Experimental Neurology 323 (2020) 113095 Contents lists available at ScienceDirect Experimental Neurology journal homepage: www.elsevier.com/locate/yexnr Research paper Helios modulates the maturation of a CA1 neuronal subpopulation required T for spatial memory formation ⁎ Albert Giralta,b,c,d, , Verónica Britob,c,d,e, Monica Pardoa,b,c,d, Sara E. Rubiod,f, Lucile Marion-Pollg,h,i, Raquel Martín-Ibáñeza,b,c,d, Alfonsa Zamora-Moratallaj, Carles Boschb,d,f, Jesús J. Ballesterosj, Esther Blascok, Aida García-Torralbal,m, Marta Pascualb,d,f, Martí Pumarolak, Jordi Alberchb,c,d,e,n, Silvia Ginésb,c,d,e, Eduardo D. Martínj, Jose Segovial,m, ⁎ Eduardo Sorianob,d,f,o, Josep M. Canalsa,b,c,d,n, a Stem Cell and Regenerative Medicine Laboratory, Department of Biomedical Sciences, Faculty of Medicine and Health Sciences, University of Barcelona, Barcelona, Spain b Neurosciences Institute, University of Barcelona, Barcelona, Spain c Institut d'Investigacions Biomèdiques August Pi i Sunyer (IDIBAPS), Casanova 143, Barcelona, Spain d Centro de Investigación Biomédica en Red sobre Enfermedades Neurodegenerativas (CIBERNED), Spain e Pathophysiology of Neurodegenerative Diseases Laboratory, Department of Biomedical Sciences, Faculty of Medicine and Health Science, University of Barcelona, Barcelona, Spain f Developmental Neurobiology and Regeneration Laboratory, Dept. Cell Biology, Physiology and Immunology, Faculty of Biology, University of Barcelona, Barcelona, Spain g Inserm UMR-S 839, 75005 Paris, France h Sorbonne Université, Faculté des Sciences et d'Ingénierie, 75005 Paris, France i Institut du Fer à Moulin, 75005 Paris, France j Laboratory of Neurophysiology and Synaptic Plasticity, Instituto Cajal, Consejo Superior de investigaciones Científicas, Madrid, Spain k Departament de Medicina i Cirurgia Animals, Universitat Autònoma de Barcelona, Cerdanyola del Vallès, Barcelona, Spain l Division of Hematopoietic Innovative Therapies, Centro de Investigaciones Energéticas, Medioambientales y Technológicas, Centro de Investigación en Red de Enfermedades Raras, Madrid, Spain m Advanced Therapies Unit, Insituto de Investigación Sanitaria Fundación Jiménez Díaz, Madrid, Spain n Production and Validation Center of Advanced Therapies (Creatio), Faculty of Medicine and Health Science, University of Barcelona, Barcelona, Spain o ICREA Academia, Barcelona, Spain A R T I C LE I N FO A B S T R A C T Keywords: Currently, molecular, electrophysiological and structural studies delineate several neural subtypes in the hip- Long-term potentiation pocampus. However, the precise developmental mechanisms that lead to this diversity are still unknown. Here VSNL1 we show that alterations in a concrete hippocampal neuronal subpopulation during development specifically Hippocampus affect hippocampal-dependent spatial memory. We observed that the genetic deletion of the transcription factor Memory Helios in mice, which is specifically expressed in developing hippocampal calbindin-positive CA1 pyramidal Development neurons (CB-CA1-PNs), induces adult alterations affecting spatial memory. In the same mice, CA3-CA1 synaptic Dendritic spines plasticity and spine density and morphology in adult CB-CA1-PNs were severely compromised. RNAseq ex- periments in developing hippocampus identified an aberrant increase on the Visinin-like protein 1 (VSNL1) expression in the hippocampi devoid of Helios. This aberrant increase on VSNL1 levels was localized in the CB- CA1-PNs. Normalization of VSNL1 levels in CB-CA1-PNs devoid of Helios rescued their spine loss in vitro. Our study identifies a novel and specific developmental molecular pathway involved in the maturation and function of a CA1 pyramidal neuronal subtype. 1. Introduction nucleus that controls episodic memory, while other nuclei such as the striatum are involved in procedural memory (Morris and Frey, 1997; Brain circuits show selective connectivities (Brown and Hestrin, Packard and Knowlton, 2002). The extent to which individual neurons 2009; Yoshimura and Callaway, 2005) that might structure information are interconnected selectively within brain circuits is an unresolved processing and its storage. In particular, the hippocampus is the main problem in neuroscience. During the recent years, several laboratories ⁎ Corresponding authors at: Department of Biomedical Sciences, Faculty of Medicine and Health Science, University of Barcelona, Barcelona. Spain. E-mail addresses: albertgiralt@ub.edu (A. Giralt), jmcanals@ub.edu (J.M. Canals). https://doi.org/10.1016/j.expneurol.2019.113095 Received 19 June 2019; Received in revised form 17 October 2019; Accepted 29 October 2019 Available online 08 November 2019 0014-4886/ © 2019 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/BY-NC-ND/4.0/). A. Giralt, et al. Experimental Neurology 323 (2020) 113095 have elucidated and characterized several subtypes of neuronal types 2. Materials and methods that makes the view of the main tri-synaptic feed forward hippocampal circuit a bit limited. Indeed, distinct research groups have shown the 2.1. Animals rich variety of neuronal subpopulations in the separate hippocampal subfields, namely CA1, CA2, CA3, CA4 and DG (Cembrowski et al., For this study we used wild type (He+/+) and knockout He (He−/−) 2016; Deguchi et al., 2011; Dong et al., 2009; Slomianka et al., 2011; female mice (Cai et al., 2009) in a C57/BL6 strain background which Thompson et al., 2008). The topographic differences are observed from were generously provided by professor Katia Georgopoulos (Harvard rostral to caudal, from dorsal to ventral and from medial to distal. These Medical School, USA). Heterozygous mice (He+/−) were removed from differences have been shown by delineating architectonic, morpholo- the study since they were indistinguishable from He+/+ mice (data not gical, transcriptomic, proteomic and electrophysiological profiles shown). Mouse genotype was determined from a tail biopsy as de- (Cembrowski et al., 2016; Deguchi et al., 2011; Dong et al., 2009; scribed elsewhere (Martin-Ibanez et al., 2012). All mice were housed Slomianka et al., 2011; Thompson et al., 2008). To unravel how these together in numerical birth order in groups of mixed genotypes (3–5 specific neuronal subpopulations, develop, mature and establish ap- mice per cage), and data were recorded for analysis by marks in their propriate connections is even more challenging and in turn, less stu- fingers or tails. The animals were housed in specific pathogen free (SPF) died. conditions and with access to food and water ad libitum in a colony The “critical period”, in terms of spinogenesis and neuronal ma- room kept at 19–22 °C and 40–60% humidity, under an inverted turation, of the hippocampal development is very sensitive to external/ 12:12 h light/dark cycle (from 08:00 to 20:00). All animal procedures environmental as well as internal/genetic challenges capable to alter were approved by local committees [Universitat de Barcelona, CEEA the proper connectivity and maturation of the region. Furthermore, it (133/10); Generalitat de Catalunya (DAAM 5712)], in accordance with has been shown that such changes cannot be observable until the brain the European Communities Council Directive (86/609/EU). reaches the maturity (Huang et al., 2014; Li et al., 2015). One of the most sensitive processes during hippocampal development is spino- 2.2. Characterization of neonatal behavior genesis. Dendritic spine number undergoes substantial changes during development, and abnormalities in these processes likely contribute to He+/+ and He−/− female pups were assessed using a slightly the pathology of several brain disorders. Spine pathology can affect modified Fox battery to study reflex and sensorimotor ontogeny both neurodevelopmental and neurodegenerative diseases and it is (Dierssen et al., 2002; Fox, 1965; Venerosi et al., 2001). The tests were widely thought that these alterations are the intrinsic cause of the ac- conducted during the dark period between 9 AM and 2 PM under dim companying neurological disturbances (Fiala et al., 2002). Some mo- light conditions. Each subject was tested at approximately the same lecules that could be involved with these processes are the calcium time of the day from postnatal day P1 to P14 depending on the test. sensors like for example the Visinin like-1 (VSNL1) protein. Indeed, During testing, animals were removed from their home cages and mutations in the VNSL1 gene have been linked to hippocampal-related maintained at constant warm temperature by a small electric mat. Body neurological diseases such as Alzheimer's disease and schizophrenia weight and survival rate were also recorded. The following reflexes and (Braunewell et al., 2011a, 2011b; Hollingworth et al., 2012). responses were scored: Grasping reflex (if the mouse displayed the re- Ikaros and Helios, which are members of the same family, have been flex was quantified as 0 = yes, or 1 = not), righting reflex (seconds that described as transcription factors with zinc-finger domains (Cobb and the pup took to rightly place its body after putting it lying face up), cliff Smale, 2005). We have previously reported the expression of Ikaros drop aversion (seconds that the pup took to remove the forepaws from (Martin-Ibanez et al., 2010) and Helios (Martin-Ibanez et al., 2012) in the cliff and turned away 90° from it), negative geotaxis test (seconds the central nervous system. Ikaros expression is mainly detected in the that the pup took to turn 180° and started to climb), clinging and embryonic lateral ganglion eminence (striatum) and postnatal cere- climbing (seconds in which the pup was able to remain in the wire bellum (Martin-Ibanez et al., 2010), and its expression is coincident (moving or not), locomotor activity and “pivoting” (number of 2.5 cm2 with its role in the development of striatal projecting neurons (Agoston squares crossed with the four paws and number of 180° turns respec- et al., 2007; Martin-Ibanez et al., 2010). Interestingly, Helios has been tively). detected in the hippocampal CA1 and the striatum at embryonic but not adult stages (Martin-Ibanez et al., 2012), which suggests a role for this 2.3. SHIRPA assessment transcription factor in the developmental regulation of neural systems involved in intellectual functions. In particular, the constricted Helios To provide detailed general phenotype assessment, mice were tested expression in the CA1 suggests that this transcription factor could play a using a modified version of the primary screen so-called Phenotype relevant role in the modulation of cognitive functions dependent on this Assessment (SHIRPA) protocol. Briefly, we measured several para- hippocampal region such as episodic memory (Sellami et al., 2017; meters such as body weight (g) and body length (cm), body position Dimsdale-Zucker et al., 2018; Stevenson et al., 2018). (curved or normal), tremors (present or not), palpebral closure (eyes In the present work, we examined how lack of Helios expression visible or visually closed), coat appearance (presence or not of piloer- affects hippocampal development and mouse behavior during neonatal ection), number of whiskers (present or clearly reduced), defecations development and in the adulthood. Our results demonstrated that adult (presence of defecations or not), locomotor activity (number of 2.5 cm2 mice devoid of Helios (He−/−) display specific hippocampal-related squares crossed during 30 s), touch escape reflex (animal reacts or not cognitive deficits, whereas neonatal He−/− mice behaved normally. to touching), corneal reflex (presence or not of palpebral closure reflex Biochemical and synaptic plasticity experiments carried out in adult when touching), biting behavior (animal bites or not a pencil when He−/− mice demonstrated that Helios has a specific and important role placing it close to its mouth), vocalization (the mouse screams or not) for a correct function of hippocampal calbindin-positive CA1 pyramidal and muscular strength (animal keeps at least 30 s in a turned wire or neurons regulating spinogenesis and, furthermore, in vitro experiments not). For further information, the instructions and the full list of para- indicated that Visinin-like protein 1 (VSNL1) could play an important meters analyzed and how are they analyzed are provided by the role in such processes. Medical Council Research Harwell at the website: http://www.mgu.har. mrc.ac.uk/MGU-welcome.html. 2.4. Visual placing response and light/dark choice test The visual placing response was tested in order to evaluate the 2 A. Giralt, et al. Experimental Neurology 323 (2020) 113095 function of the visual system (Metz and Schwab, 2004). For the light/ which the platform was switched to the left arm of the T-maze. Then dark choice test we used a corridor made of black perspex (40 cm long x mice received 12 additional trials divided in 3 blocks of 4 trials. 10 cm wide x 20 cm high) with one well-lighted side (w60 W lamp; 1000 lx) and with the other side in darkness (enclosed; 50 lx). We 2.9. T-maze spontaneous alternation (T-SAT) placed mice individually in the well-lighted side, close to the wall, and recorded the latency to reach the dark side. The procedure was carried The transparent T-maze apparatus used for the T-SAT was the same out for 3 trials with an inter-trial interval of 15 min for each mouse. used in the Strategy shifting in a water T-maze paradigm but without water. In the training trial, one arm was closed (novel arm) and mice 2.5. Open field were placed in the stem arm of the T (home arm) and allowed to ex- plore this arm and the other available arm (familiar arm) for 10 min, To check spontaneous locomotor activity and habituation we used after which they were returned to the home cage. After inter-trial in- the open field (Giralt et al., 2010). Briefly, the apparatus consisted of a terval of 2 h mice were placed in the stem arm of the T-maze and al- white square arena measuring 40 × 40 × 40 cm. Light intensity was lowed to freely explore all three arms for 5 min (testing phase). Because ~150 lx throughout the arena. Animals were placed on the arena center the T-maze was made of transparent methacrylate (0.3 cm of thickness), and allowed to explore freely for 15 min during three consecutive days. 5 objects of similar size (~20 cm3) and highly perceptible were situated Time in the center and pathlength (cm) were measured as measures of surrounding the maze at ~15–20 cm outside the walls. The first choice anxiety-like behaviors and locomotor activity respectively. The beha- to turn either to the familiar arm or to the new arm (alternation rate, %) vior was recorded using a video camera and analyzed using the Smart and the number of arm entries (number of entries in the new arm*100/ Junior software (Panlab, Spain). total arm entries) were the two parameters evaluated in the testing phase. 2.6. Dark-light paradigm 2.10. Place and response learning in a T-maze The dark-light box consisted in a home-made box with two com- partments (Giralt et al., 2013; Brito et al., 2014), which were connected Using the same transparent T-maze and the same extra-maze visual by a door (5 cm × 5 cm) located at floor level in the center of the cues (see 2.9.) we carried out a spatial vs response learning protocol as partition (Bourin and Hascoet, 2003). The light compartment described previously (Lex et al., 2011) in a new cohort of mice to avoid (15 × 12 × 25 cm) was white and directly illuminated with light effects of previous experience with the apparatus and procedure. This (200 lx), while the dark compartment (12 × 12 × 25 cm) was black and new cohort of mice was food deprived to 85% of their normal body completely enclosed (20 lx). The test was performed in a dimly lighted weight. The food deprivation was initiated 7 days prior to the T-maze room. Each animal was initially confined in the dark compartment and experiment. During the entire experimental procedure mice had food allowed to freely explore the apparatus for 5 min. A video camera re- accessibility 1.5 h/day (1 h after the experimental session). Then the corded the animal behavior. Time to step-through for the first time and animals were habituated to the T-maze for a 5 min period. In the total time in the bright compartment were the variables evaluated and training, mice were then placed in the start arm and trained to run to analyzed using the Smart Junior software (Panlab, Spain). the end of one forced-goal arm during 4 consecutive trials with a 1 min inter-trial interval. A sweet pellet (Bio-serve, Frenchtown, NJ, USA) was 2.7. Object recognition test placed at the end of this goal-arm (baited-arm). Mice were allowed to collect and eat the sweet pellet and then returned to their home cage. Novel object recognition test (NORT) was used to analyze recogni- Latencies to reach the pellet were monitored to evaluate learning skills. tion memory as described previously. The device consisted in a white In the test, animals were placed in the arm opposite to the baited-arm circular arena with 40 cm diameter and 40 cm high. The light intensity and all arms remained opened. The first arm visited was registered to was 100 lx throughout the arena. Mice were first habituated to the evaluate the chosen strategy (response strategy: turn left; place arena in the absence of objects (3 days, 15 min/day). On the fourth day, strategy: visit the previously baited arm). two similar objects were presented to each mouse during 10 min (A'A" condition) after which they were returned to their home cage for 24 h. 2.11. RNA deep sequencing After that, the animals were placed in the arena where they were tested during 5 min with a familiar and a new object (A'B condition), and then RNAseq procedures were carried out in collaboration with IMAGIF returned to the home cage. Animal behavior was video tracked and (Yves sur Yvette, France). Briefly, library preparation was done using analyzed using the Smart Junior software (Panlab, Spain). Exploring the TruSeq mRNA Stranded library prep kit from Illumina following the behavior was considered when the animal orientated its nose rubbing/ manufacturer's instructions. Libraries were quality tested using the touching the object. The object preference was measured as the time Bioanalyzer system from Agilent and semi-quantitative PCR. exploring each object x 100/time exploring both objects. Sequencing was done using the NextSeq500 from Illumina, generating 75 bp single reads (NextSeq 500 High Output Kit 75 cycles). The data 2.8. Strategy shifting in a water T-maze were demultiplexed using the distribution of CASAVA software (CASAVA-1.8.2 bcl2fastq2 v2.15.0). The quality of the data was To evaluate the procedural learning and cognitive flexibility in mice checked with the software FastQC 0.11. Illumina adapters were re- we developed a swimming T-maze test as previously described (Van moved using Cutadapt-1.3, keeping only reads with a minimal length of Raamsdonk et al., 2005). In this test, mice were placed in the base of a 10 nucleotides. After Cutadapt step, the data were mapped using bwa- transparent water-filled T-maze with an escape submerged platform 0.7.4 on reference genome mouse mm10. The bam output files were located in the right arm of the maze (T-maze dimensions: arms, sorted with samtools. The counting of number reads in each exon has 35 × 20 cm; water depth, 10 cm; platform, 8 × 8 cm). The water tem- been done with HTSeq-count and DESeq2 for the differential analysis. perature was maintained at 26 °C ± 1. A black screen surrounded the The accession number for the RNA-seq data reported here is GEO: tank to avoid the vision of distal cues. The time and the first path taken (GSE124641). to reach the platform were recorded. The escape platform location was counterbalanced and placed in the right arm for the 50% of mice and in 2.12. Histopathological analysis the left arm for the other 50% of mice. Mice received 12 trials divided in 3 blocks of 4 trials. After 12 trials we started a reversal phase in Many tissues and organs samples were collected from He+/+ and 3 A. Giralt, et al. Experimental Neurology 323 (2020) 113095 He−/− mice at 7 weeks of age. Tissues were fixed by immersion in 10% subjected to immunofluorescence (see below) against calbindin 28 kDa neutral buffered formalin (NaCl 138 mM, KCl 2.7 mM, KH2PO4 (#300; Swant, Switzerland), incubated with DAPI (#D1306, In- 1.2 mM, Na2HPO4 8.1 mM in distilled water) for 48 h, dehydrated and vitrogen), and mounted in Mowiol to be analyzed. Spine density embedded in paraffin for the histopathologic evaluation. counting and morphology analysis was performed only in calbindin Decalcification in 10% formaldehyde, 8% plus formic acid and 1% 28 kDa positive pyramidal neurons by using the freeware Neuronstudio methanol was done in order to make the bone tissue able to be cut in (Computational Neurobiology and Imaging Center, Icahn School of the microtome. Tissue sections (3 μm) were obtained with a sliding Medicine at Mount Sinai) as previously described (Jammalamadaka microtome. Dewaxed sections were stained with hematoxylin and eosin et al., 2013). for morphological examination. 2.16. Immunohistochemistry and immunofluorescence 2.13. Golgi staining and spine counting For the immunohistochemistry experiments P0 and P5 pups were For the Golgi-Colonnier staining, adult He+/+ and He−/− mice (7- used. Animals were deeply anaesthetized and subsequently in- week-old) were perfused with 2% paraformaldehyde - 2% glutar- tracardially perfused with 4% paraformaldehyde in 0.1 M phosphate aldehyde in 0.12 M phosphate buffer. Brains were postfixed overnight. buffer. The brains were dissected out and cryoprotected with 30% su- Pieces containing the whole hippocampal formation were dissected out crose in PBS with 0.02% sodium azide and frozen in dry-ice cooled and incubated in a dichromate-Colonnier solution (3% K2Cr2O7, 5% isopentane at −80 °C. Coronal sections (30 μm) were obtained. After glutaraldehyde in H2O) for 5 days at 15 °C. Pieces were then transferred blocking, sections were incubated overnight with specific goat antibody to a 0.75% AgNO3 solution for 3 days. After embedding the pieces in against Helios (1:50; #sc-9864, Santa Cruz Biotechnology, Santa Cruz, paraffin, 150 μm thick sections were obtained using a vibratome (Leica, CA, USA). Primary antibody was visualized by sequential incubation VT1000) and subsequently dehydrated. The sections were then with biotinylated secondary antibodies (1:200, Vector Labs, mounted onto slides and coverslipped with an Araldite solution. Images Burlingame, CA, USA) and the streptavidin-peroxidase complex (1:400, were acquired with an upright microscope (Nikon E600; Nikon. Amersham Biosciences, Pittsburgh, PA; USA). The peroxidase reaction Tallahassee, Florida, USA) and processed with ImageJ software was developed with diaminobenzidine (DAB) intensified with nickel (Rasband, W.S., ImageJ, National Institutes of Health, Bethesda, ammonium sulfate and cobalt chloride (DAB/Ni-Co), and H2O2. The Maryland, USA). The density of dendritic spines was quantified in the sections were mounted onto gelatinized slides, dehydrated and cover- stratum radiatum, i.e. in the primary side branches of the apical den- slipped with Eukkit (Panreac). drite of pyramidal neurons located in the CA1 area of dorsal hippo- For double immunofluorescence studies sections from P0 mice were campi of Golgi-stained sections from He−/− and He+/+ mice (2–4 side used. Sections were incubated overnight with the specific goat antibody branches per neuron; 4–12 neurons per mice; 4 He+/+ and 3 He−/− detecting Helios (1:50) combined with one of the following antibodies: mice per genotype). Spines were quantified at 100× magnification in a mouse anti-MAP2 (1:500; #M1406, Sigma-Aldrich, St. Louis, MO, 10-μm long segment of the side branch considered. USA), rabbit anti- calbindin 28 kDa (1:5000; #CB300, Swant Antibodies, Switzerland), rabbit anti-nestin (1:1000; generous gift from 2.14. Electron microscopy S. Hockfield) and mouse anti-Ki67 (1:50; #ab16667; Abcam, Cambridge, UK). After incubation with appropriate fluorescent sec- For the electron microscopy analysis adult, He−/− and He+/+ mice ondary antibodies, nuclei were stained with 4′,6-diamidino-2-pheny- (7-week-old) were perfused with 2% glutaraldehyde 2%, paraf- lindole (DAPI; Sigma-Aldrich). Images were acquired under an SP5 ormaldehyde in fresh 0.12 M phosphate buffer. Brains were postfixed in confocal microscope (Leica Microsystems, Heidelberg, Germany), and the same fixative solution overnight. Tissue slices (150 μm thick sec- processed with ImageJ software. tions) obtained in a vibratome (Leica VT100) and containing dorsal For single immunofluorescence studies brains were removed at hippocampi were then incubated with 2% osmium tetroxide, stained specific developmental stages (from E16.5 to P35), and frozen in dry ice with 2% uranyl acetate, dehydrated and embedded in Araldite. cooled isopentane. Serial coronal cryostat sections (14 μm) were cut on Ultrathin sections were collected using a ultramicrotome (Ultracut E a cryostat, collected on silane-coated slides and frozen at −20 °C. Leica) and cut with a diamond knife on formvar-coated slot grids and Tissue sections were blocked for 1 h in PBS containing 0.3% Triton X- stained with lead citrate. Electron micrographs depicting an 18.8 μm2 100 and 1% bovine serum albumin (BSA). Thereafter, they were in- area were randomly taken in stratum radiatum layer (n = 41 micro- cubated overnight at 4 °C in PBS containing 0.3% Triton X-100 and 1% graphs per group). The number of asymmetric (Gray Type 1) synapses BSA with the antibody anti-Helios (1:50; Santa Cruz). After three PBS per field was calculated. washes, sections were incubated for 2 h at room temperature with the secondary antibodies Cy3 donkey anti-goat (1:200; from Jackson 2.15. DioListic staining and spine counting and morphology analysis Immunoresearch Laboratories Inc. Sacramento, California, USA). At the end, tissue sections were counterstained with DAPI. Images were ac- Hippocampal neurons were labeled using the Helios Gene Gun quired under an SP5 confocal microscope. System (Bio-Rad) as previously described (Brito et al., 2014). Briefly, a suspension buffer containing 3 mg of DiI (Molecular Probes, Invitrogen) 2.17. Confocal imaging dissolved in 100 μl of methylene chloride (Sigma-Aldrich) and mixed with 50 mg of tungsten particles (1.7 mm diameter; Bio-Rad) was DiI-labeled pyramidal neurons from CA1 of the dorsal hippocampus spread on a glass slide and air-dried. The mixture was resuspended in and medium spiny neurons from the dorsal striatum were imaged using 3.5 ml distilled water and sonicated. Subsequently, the mixture was a Leica Confocal SP5 with a × 63 oil-immersion objective (digital zoom drawn into Tefzel tubing (Bio-Rad), and then removed to allow tube 5×) as previously described (Brito et al., 2014). In vitro confocal ima- drying during 5 min under a nitrogen flow gas. Then, the tube was cut ging were obtained as for common images as well as for spine density into 13-mm pieces to be used as gene gun cartridges. Dye-coated par- and morphology imaging as previously described (Giralt et al., 2017). ticles were delivered in the hippocampus using the following protocol. Shooting was performed over 200-μm coronal sections (obtained in a 2.18. Electrophysiology vibratome, Leica VT100) at 80 psi through a membrane filter of 3 μm pore size and 8 × 10 pores/cm2 (Millipore). Sections were stored at Transverse brain slices (400 μm thickness) obtained using a vi- room temperature in PBS for 3 h protected from light and then bratome (Leica, VT1000) were prepared from He+/+ and He−/− mice 4 A. Giralt, et al. Experimental Neurology 323 (2020) 113095 (7-week-old). Slices were transferred to an immersion recording constructs transfection, hippocampal neurons at DIV 21 were trans- chamber and superfused (2.5 ml/min) with equilibrated ACSF main- fected using transfectin (Bio-Rad, Hercules, CA, USA) following the tained at constant temperature (34 °C). Evoked fEPSPs were elicited by manufacturer's instructions and left for 72 h. Cells were co-transfected stimulation of the Schaeffer collateral fibers with an extracellular bi- with a GFP-expressing construct (Giralt et al., 2017) together with a polar tungsten electrode via a 2100 isolated pulse stimulator (A-M mCherry-shRNA-Scramble or mCherry-shRNA-VSNL1 construct. Systems, Inc., Carlsborg, WA, USA). The stimulation intensity was ad- justed to give fEPSP amplitude that was approximately 50% of maximal 2.22. Experimental design and statistical analysis fEPSP sizes. Long-term potentiation (LTP) was induced by applying 4 trains (1 s at 100 Hz) spaced 20 s, and potentiation was measured for All data are expressed as mean ± SEM. Statistical analysis were 1 h after LTP induction at 0.1 Hz. Electrophysiological cortico-striatal performed by using the unpaired Student's t-test (95% confidence), one- experiments were performed as described elsewhere (Giralt et al., way and two-way ANOVA with Tukey as a post hoc test. Fisher's exact 2011a). LTP was induced by applying four 1 s, 100 Hz trains delivered test, Chi2 (χ2) test and logistic regression were also used as appropriate every 10 s, and potentiation was measured for 1 h after LTP induction at and indicated in the figure legends. Values of p < .05 were considered 0.1 Hz. For each experiment, fEPSP slopes or PS amplitude were ex- as statistically significant. pressed as a percentage of average pre-tetanus baseline slope or am- All experiments in this study were blinded and randomized. All mice plitude values respectively. bred for the experiments were used for preplanned experiments and randomized to experimental groups. Visibly sick animals were excluded 2.19. Western blotting before data collection and analysis. Data were collected, processed and analyzed randomly. The experimental design and handling of mice For western blot analysis we first dissected out the dorsal hippo- were identical across experiments. Littermates were used as controls campal region bilaterally from adult (8-week-old) or neonatal (P8) with multiple litters (3–5) examined per experiments. All mice were He+/+ and He−/− mice. Hippocampal homogenates processing and bred in the Animal Facility of the Medical School in the University of western blot protocol were carried out as described elsewhere (Giralt Barcelona. et al., 2011a). Immunoblots were probed with the following primary antibodies: VNSL1 (Proteintech, #13919–1-AP, Manchester, UK), anti- 3. Results GluN2A (Millipore, #AB1555, Darmstadt, Germany), anti-GluN1 (Chemicon, #MAB363, Temecula, CA, USA), anti-GluN2B (Chemicon, 3.1. Helios is localized in developing calbindin-positive pyramidal neurons #AB1557); anti-PSD-95/SAP-90 (Cell Signaling, #3450S, Leiden, Nee- of the CA1 therlands), anti-SynGAP (Affinity BioReagents, #PA1–046, Golden, CO, USA); anti-Chapsyn110 (Alomone Labs, #APZ-002, Jerusalem, Israel), Helios has been recently described to be expressed in ectodermal anti-CaM Kinase II (Santa Cruz Biotechnology, #sc-9035 (M-176), and neuro-ectodermal derived tissues (Martin-Ibanez et al., 2012). To Santa Cruz, CA, USA), and anti-GluN3A (was a gift from Dr. Isabel Pérez further analyze the expression pattern of this transcription factor in the Otaño, Alicante, Spain). After primary antibody incubation, blots were brain, we performed a broad immunofluorescence characterization at incubated with anti-mouse or anti-rabbit IgG HRP-conjugated (Pro- several developmental stages: E14.5, E16.5, E18.5, P2, P7, P15 and mega, Madison, WI, USA) and developed using the ECL Western blot- P35. We observed that Helios expression is mostly localized in the ting analysis system (Santa Cruz Biotechnology). A monoclonal anti-β- pyramidal layer of the hippocampal CA1, the striatum and the Purkinje tubulin antibody (Sigma, #083M4847V) was used to obtain loading cell layer of the cerebellum (Supplementary Table 1). Within most of controls. The intensity of the optical density of the bands was quantified these brain areas, Helios presented a dynamic and transient expression using the GelPro 3.2 software (Infaimon). pattern starting at embryonic developmental stages (E14.5-E16.5), peaking between E18.5-P3 and disappearing after P15 (Supplementary 2.20. Hippocampal primary cultures Table 1). However, in the cerebellum Helios started to be expressed at P7 and lowered its expression by P35, according to the delayed post- Hippocampal neurons were prepared at E17 from our Helios (+/+ natal development of Purkinje cells. A common feature for all the brain or −/− pups) colony as described elsewhere (Giralt et al., 2017). At regions analyzed was the absence of Helios expression in the adulthood. DIV 20–22, cells were transfected (see below) being fixed 72 h later for We next further characterized Helios expression and localization in the 10 min with 40 g l−1 paraformaldehyde in PB 0.2 M for immunostaining. hippocampus during development. At E16.5, Helios expression started Fixed cells were permeabilized in 1 ml l−1 Triton X-100 for 10 min and to be evident in cells of the hippocampal plate of the CA1 (Fig. 1A-B). then blocking was performed with 10 g l−1 BSA in PBS for 1 h. Cells These Helios-positive cells were negative for both, Ki67 (Fig. 1C) and were incubated with mouse monoclonal antibody for calbindin, nestin (Fig. 1D). At P0, Helios immunostaining still labeled cells in the (1:1000, Swant, Switzerland) at 4 °C overnight. After three washes with stratum pyramidale of CA1 (Fig. 1E-F). In addition, some cells in the PBS, cells were incubated with the corresponding fluorescent secondary subiculum also expressed Helios at this stage (Fig. 1E-F). No Helios antibody Alexa anti-mouse 647 (1:250; ThermoFisher, UK). After expression was observed in either CA3 or DG (Fig. 1B and E), indicating washing twice with PBS, the coverslips were mounted with Vectashield a very specific localization within the hippocampus. (Vector Laboratories Burlingam, UK). Hippocampal neuron staining Double immunostaining for MAP2 and Helios at P0 showed that all was observed with a confocal SP5 (see above). Helios-immunoreactive cells were immunostained with the neuronal marker MAP2 in the pyramidal layer of CA1 (Fig. 1G). All Helios-po- 2.21. Construct generation and transfection sitive cells displayed the typical shape of pyramidal-like neurons. To further characterize Helios-positive pyramidal neurons, we used the Five different shRNA were designed along Visinin-like 1 mRNA, that calbindin-28 kDa (CB) marker. Our results showed that virtually none consist of sense and antisense sequences (19–21 nt in length) separated of the CB-positive neurons in CA1 displaying morphology of inter- by a loop structure (TTCAAGAGA) and a 5′ AAAA overhang. Detailed neurons were positive for Helios (Fig. 1H, arrows). Conversely, we sequences of the shRNAs are described in Fig. 7A. The sequences in- could observe that most of the CB-positive cells in the CA1 pyramidal cluded additional restriction sites (MluI in 5′ and XhoI –ClaI in 3′) to layer co-expressed the Helios protein (Fig. 1H arrowhead). These Helios facilitate their cloning into pLVTHM vector (kindly provided by Dr. positive neurons exhibited pyramidal-like shapes. Thus, our data show Trono). Expression of the shRNAs is driven by H1 promoter and that Helios is transiently expressed in the CB-positive subpopulation of mCherry expression, as a marker, is driven by EF1-alpha promoter. For pyramidal neurons (CB-CA1-PNs) in the CA1 of the hippocampus. 5 A. Giralt, et al. Experimental Neurology 323 (2020) 113095 (caption on next page) 6 A. Giralt, et al. Experimental Neurology 323 (2020) 113095 Fig. 1. Helios expression pattern localized in the CA1 hippocampal plate/stratum pyramidale during development. (A) Double staining for DAPI (blue) and Helios (red) in the hippocampus. From left panel to right panel: embryonic days 16.5 and 18.5 and postnatal days 2 and 7. Scale bar: 70 μm. (B) General view of the Helios expression pattern (green) in the hippocampal formation at embryonic days 16.5. Scale bar: 400 μm. (C) Double staining for Helios (red) and Ki67 (green) in the CA1 at embryonic day 16.5. Scale bar: 40 μm. (D) Double staining for Helios (red) and Nestin (green) in the CA1 at embryonic day 16.5. (E) Immunohistochemistry for Helios (DAB method) in the hippocampal formation at postnatal day 0. Scale bar: 400 μm. (F) Inset of the immunohistochemistry for Helios (DAB method) in the CA1 of the hippocampal formation at postnatal day 0. Scale bar: 70 μm. (G) Double staining for Helios (green) and MAP2 (red) in the CA1 at postnatal day 0. Scale bar: 50 μm. Inset scale bar: 30 μm. (H) Double staining for Helios (green) and calbindin-D28k (red) in the CA1 at postnatal day 0. Scale bar: 50 μm. Scale bar inset: 15 μm. IZ: Intermediate zone; HP: Hippocampal plate; SO: Stratum oriens; SP: Stratum pyramidale; SR: Stratum radiatum; SU: Subiculum; DG: Dentate Gyrus; CA1-CA3: Cornu amonis 1 and 3; IBH: Hippocampal inferior band; CB: Calbindin-D28k. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) Fig. 2. Neurological and reflexive characterization of He−/− pups during postnatal development. (A) Latencies in the righting reflex test were monitored in He+/+ and He−/− pups from P1 to P8 (genotype effect: F(1, 15) = 1.306, p = n.s.). (B) The latency to show the response to negative geotaxis was evaluated in He+/+ and He−/− pups from P1 to P5 (genotype effect: F(1, 15) = 6.614, p = .0121). (C) Aversion to falling from a cliff was checked by measuring the latency to avoid the cliff in He+/+ and He−/− pups from P2 to P4 (genotype effect: F(1, 15) = 0.1346, p = n.s.). (D) Muscular strength and tone were evaluated by the clinging/climbing response and latency to fall from a 90°-placed grid recorded in He+/+ and He−/− pups from P8 to P14 (genotype effect: F(1, 15) = 3.362, p = n.s.). (E-F) The development of motor responses and walking maturation were evaluated in an open field (20 cm2) in He+/+ and He−/− pups from P3 to P14. We recorded the number of 2.5 × 2.5 cm crossed squares (E; genotype effect: F(1, 15) = 0.3870, p = n.s.) and number of 90° pivoting movements (F; genotype effect: F(1, 15) = 0.0013, p = n.s.). Values are expressed as mean ± SEM. Data was analyzed by two-way ANOVA with repeated measures in all cases. * p < .05 as compared to He+/+ mice (n = 9 He+/+; 8 He−/−). 3.2. Normal neurological state in neonatal He−/− mice development was carried out in He+/+ and He−/− pups. First, He+/+ and He−/− pups were able to exhibit a progressive and faster righting In order to study the role of Helios during development a compre- reflex response from post-natal day 1 (P1) to post-natal day 8 (P8) hensive neurobehavioral characterization during the postnatal without significant differences between genotypes (Fig. 2A). Next, we 7 A. Giralt, et al. Experimental Neurology 323 (2020) 113095 evaluated negative geotaxis from P3 to P7. He+/+ and He−/− mice after training explored with greater preference the new object without were able to elicit a proper progressive and faster response at all ages significant differences between genotypes, indicating that object re- tested (Fig. 2B). Interestingly, He−/− mice took even shorter time re- cognition memory was not impaired in He−/− mice (Fig. 3A). Next, actions than He+/+ mice to show appropriate turning-and-walking He+/+ and He−/− mice were subjected to the strategy-shifting task in a behavior in this paradigm (Fig. 2B). In addition, He+/+ and He−/− swimming T-maze (Fig. 3B-C). First, mice were trained to swim to reach pups were able to successfully demonstrate a cliff aversion response the submerged platform located in one arm. In this acquisition phase of from P2 to P4 (Fig. 2C). We next performed a sensorimotor character- the task He+/+ and He−/− mice rapidly learned to swim to the plat- ization in pups to check maturation of sensory and motor systems and form without significant differences between genotypes, either in the muscular strength. For the clinging/climbing test, all pups progressively time taken to reach the platform (Fig. 3B), or in the first path to reach improved the time clinging in the wire mesh from P8 to P14 (Fig. 2D). the platform (Fig. 3C). Switching the location of the platform to the Additionally, all animals grabbed successfully to the grid (data not opposite arm of the maze in the reversal phase revealed no difference shown). Furthermore, the hanging time was not significantly different between He+/+ and He−/− mice either in time taken to reach the between He−/− and He+/+ mice (Fig. 2D). Pivoting behavior was platform (Fig. 3B) or in the first path to reach the platform (Fig. 3C). significantly decreased with age from P3 to P14 (Fig. 2E). In an op- These results indicate that He−/− mice did not display alterations in posing trend, spontaneous motor activity was significantly increased the strategy shifting ability. In the T-SAT task, spontaneous alternation with age in both genotypes (Fig. 2F), indicating a normal progressive and arm preference were evaluated 2 h after habituation. We found that rostro-caudal maturation of the limbs in both, He+/+ and He−/− mice. He+/+ mice significantly alternated whereas He−/− mice did not Neither pivoting behavior (Fig. 2E) nor spontaneous locomotor activity (Fig. 3D). Furthermore, He+/+ mice visited significantly more times (Fig. 2F) were significantly different between genotypes. Altogether the novel arm than the familiar arm whereas He−/− mice showed no indicate that neurological and reflexive behavior in He−/− pups were preference for any arm (Fig. 3E). Thus, in the T-SAT He−/− mice almost indistinguishable from those on He+/+ pups. showed alterations in both variables, spontaneous alternation and new- context exploration. Finally, to evaluate whether He−/− mice display 3.3. Specific spatial learning and memory deficits in adult he−/− mice aberrant strategies in the spontaneous alternation in a T-maze, we used a protocol to detect likely aberrant strategies to solve this task (Lex After the broad neonatal characterization, we next performed a et al., 2011). A new cohort of animals was habituated to the T-maze for comprehensive behavioral characterization in young He+/+ and He−/ a 5 min period with the start arm and one of the lateral arms opened. − mice from 5 to 7 weeks of age (See Supplementary fig. 1 for the se- Both genotypes explored similarly the maze (data not shown). Then, the quence of the behavioral tests). First, the animals were evaluated by the training procedure started and latencies to collect and eat the sweet SHIRPA protocol (at 5 weeks of age) which it gives indications about pellet placed at the end of the baited arm were similar between groups specific behavioral anomalies resulting from muscular, lower motor (data not shown), suggesting similar levels of performance to reach the neuronal, spinocerebellar, sensory, neuropsychiatric, and autonomic sweet pellet. Finally, we found that He+/+ mice displayed a place defects. As a result (Supplementary Table 2), He−/− mice showed learning strategy whereas He−/− mice significantly showed response shorter body length and lower body weight compared to He+/+ mice as learning (Fig. 3F). Altogether, these results indicated that spatial previously reported (Cai et al., 2009). Additionally, He−/− mice learning, but not object recognition memory, procedural learning or showed less palpebral closure (but normal palpebral reflex) and an cognitive flexibility was affected in He−/− mice. absence of startle response compared to He+/+ mice (Supplementary Table 2). We further investigated the less palpebral closure as it could 3.4. Specific impairment of synaptic plasticity in the Schaffer collateral affect ulterior behavioral analysis. Necropsy and histological analysis pathway but not in the cortico-striatal pathway in adult He−/− mice indicated that less palpebral closure in He−/− mice is due to a discrete inflammatory focus near to the sebaceous acini in the tarsal, or in the Our results suggested that He−/− mice present cognitive deficits Meibomian gland in the eyelids (Supplementary Fig. 2A). However, this related with hippocampal function, but not with cortico-striatal func- alteration in the He−/− mice eyelids did not affect either, the retina tion. Thus, we next investigated excitatory synaptic transmission and structure (Supplementary Fig. 2B) or the visual acuity (Supplementary plasticity at Schaffer collateral-CA1 hippocampal synapses and cortico- Fig. 2C-D). striatal synapses in slices from 7-week-old He+/+ and He−/− mice. Anxiety and locomotor activity and exploration were also evaluated First, we analyzed basal synaptic transmission by applying isolated in He+/+ and He−/− mice. Anxiety levels were evaluated by using the stimuli of increasing intensity to the Schaffer collaterals. For a range of light-dark paradigm. Both, the total time remaining in the light zone stimulation intensities, the slope of fEPSP responses from He−/− slices (Supplementary Fig. 2E) and the time to cross from the dark zone to the was not significantly different from He+/+ slices (Fig. 4A, n = 11 light zone (Supplementary Fig. 2F) were similar in He+/+ and He−/− slices, 4 mice). Likewise, measurements of the fiber volleys from He−/− mice. The comparative exploratory activity from the open field test is and He+/+ slices were similar (Fig. 4B, n = 11 slices, 4 mice), and shown in (Supplementary Fig. 2G-I). He+/+ and He−/− mice did not there were no differences in an input/output curve (Fig. 4C). Then, we differ in their first-to-third day covered distance (Supplementary investigated synaptic plasticity at CA1 hippocampal synapses using a Fig. 2G) or in the time spent in the center (Supplementary Fig. 2H). high-frequency conditioning tetanus to induce LTP. Baseline responses Furthermore, a similar significant change in activity from day 1 to day 3 were monitored for 10–30 min before conditioning and were found to was observed in both genotypes regarding to the covered distance be stable. He−/− mice showed diminished tetanic conditioning-LTP (Supplementary Fig. 2G) and center preference (Supplementary compared to He+/+ animals (Fig. 4D). At 60 min after tetanus, po- Fig. 2H). tentiation (as mean percentage of baseline) in He+/+ mice was Next, we evaluated high cognitive function by carrying out a set of 150.4 ± 8% (n = 11 slices) versus 121.5 ± 5% in He−/− mice distinct learning and memory paradigms in young (~8 weeks of age) (Fig. 4D; p < .01, n = 12 slices). Thus, these data represent a possible He+/+ and He−/− mice. We performed the novel object recognition cellular mechanism for the specific hippocampal-related cognitive task (NORT), the T-maze spatial alternation task (T-SAT) and the deficits observed in He−/− mice. strategy shifting in a water T-maze involving different dimensions of Next, we examined glutamatergic synaptic transmission in cortico- the mouse cognition. In the NORT, animals were subjected to a training striatal synapses. Extracellular recordings of population spikes (PS) in session in the arena in the presence of two similar objects (A and A′). acute cortico-striatal brain slices reflect excitatory monosynaptic He+/+ and He−/− mice similarly explored the object A and A′ in- transmission in a population of striatal neurons due to endogenous dicating no object or place preferences (Fig. 3A). Animals assessed 24 h glutamate release upon electrical stimulation of glutamate-containing 8 A. Giralt, et al. Experimental Neurology 323 (2020) 113095 Fig. 3. Specific hippocampal-related learning alterations in He−/− mice. He+/+ and He−/− mice at ~8 weeks of age were subjected to different tests of learning of memory. (A) Long-term recognition memory was evaluated in the NORT. Graph shows the percentage of object preference in He+/+ and He−/− mice during training and testing (24 h after training; learning effect: F(1, 32) = 50.22, p < .001; genotype effect: F(1, 32) = 2.248e-005, p = .9962). (BeC) Strategy-shifting capacity was assessed in a swimming T-maze in He+/+ and He−/− mice. Latency to reach the platform (B) and path to reach platform (C) were monitored. (D-E) Spatial spontaneous alternation learning was assessed by the T-SAT in He+/+ and He−/− mice. Animals were tested for spontaneous alternation rate (D; z = 10.38; p < .001) and number of arm entries (E; learning effect: F(1, 38) = 31.56, p < .001; genotype effect: F(1, 38) = 19.65, p < .001) 2 h after the training trial. (F) Place- vs Response-learning were tested in a new set of He+/+ and He−/− mice in the same T-maze apparatus (z = 10.95; p < .001). Bars represent mean ± SEM. Data were analyzed by one-way ANOVA with Tukey's t-test as a post hoc (A and E), by the Chi-square (χ2) test (D and F), by two-way ANOVA with repeated measures (B) and by logistic regression with the Wald test as a post hoc (C). In A and E: ** p < .01, *** p < .001 when compared to time exploring object A / number of visits of the old arm. In D and F: *** p < .001 when compared to He+/+ mice. n = 9 He+/+ and 8 He−/− in A-C; n = 12 He+/+ and 9 He−/− in D and E; n = 8/genotype in F. fibers. We analyzed basal synaptic transmission by applying isolated and He+/+ mice have similar hippocampal volume (Supplementary stimuli of increasing intensity. Extracellular field recordings showed Fig. 3A-B) and density of CA1 pyramidal cells (Supplementary Fig. 3C). that amplitude of PS responses in He−/− mice was not significantly Our next goal was to evaluate possible microstructural changes by different from PS response in He+/+ mice (Fig. 4E). Finally, we also analyzing neuronal morphology and by counting dendritic spines and investigated cortico-striatal synaptic plasticity by inducing LTP by high- synaptic contacts in the stratum radiatum of the CA1 in adult He−/− and frequency conditioning tetanus. Baseline responses were monitored for He+/+ mice. We only analyzed symmetric synapses and dendritic 10 min before conditioning and were found to be stable. Tetanus con- spines because Helios is specifically expressed in a subset of pyramidal ditioning revealed a similar ability to support LTP in He−/− and He+/+ neurons of the CA1 and because these micro-structures have been more mice (Fig. 4F). Taken together, these data indicate that there is a strong widely characterized in the literature about the modulation of LTP and correlation between the specific hippocampal-dependent cognitive spatial memory than symmetric/inhibitory synapses (Bourne and deficits observed in Fig. 3 and the specific impairments on hippocampal Harris, 2011). First, we performed a Golgi staining and observed no CA3-CA1 LTP. gross changes in morphology of CA1-pyramidal neurons in He−/− mice compared to He+/+ mice (Supplementary Fig. 3D). Subsequent den- 3.5. Helios deficiency induces specific dendritic spine density and dritic spine quantification in 2nd-order apical dendrites from the CA1 morphology alterations in calbindin-positive CA1 pyramidal neurons pyramidal neurons (localized in the stratum radiatum) showed no sig- nificant differences in the number of these structures in He−/− mice We next evaluated whether hippocampal architecture correlated compared to He+/+ mice (Supplementary Fig. 3E-F). We also quanti- with above described hippocampal-related cognitive impairments. fied the density of excitatory synapses in the stratum radiatum by elec- First, stereological analysis revealed that adult (7–8-week-old) He−/− tron microscopy, but no significant changes were detected between 9 A. Giralt, et al. Experimental Neurology 323 (2020) 113095 Fig. 4. Characterization of basal synaptic function and synaptic plasticity in cortico-striatal and CA3-CA1 pathways in He−/− mice. (A) Hippocampal field excitatory postsynaptic potential (fEPSP) slopes in 7-week-old He+/+ (n = 11 slices) and He−/− (n = 11 slices) mice for a given range of stimulus intensities. (B) Fiber volley amplitudes in hippocampal slices from He+/+ (n = 11 slices) and He−/− (n = 11 slices) mice for a given range of stimulus intensities. (C) Input/output relationships for hippocampal slices from He+/+ (n = 11 slices) and He−/− (n = 9 slices) mice. (D) Summary data showing the time course of mean fEPSPs slope in He+/+ (n = 11) and He−/− (n = 12) mice in basal condition and following hippocampal Schaffer-collateral LTP induction. (E) Population spike (PS) amplitudes for a given range of stimulus intensities are shown from He+/+ (n = 13) and He−/− (n = 9) cortico-striatal slices. (F) Summary data showing the time course of mean PS amplitude in cortico-striatal slices from He+/+ (n = 7) and He−/− (n = 6) mice in basal condition (10 min) and following LTP induction (60 min). Data were normalized for each slice with respect to the average slope recorded during baseline. Bars represent mean ± SEM. For statistical analysis two-way ANOVA was used. ** p < .01 with respect to He+/+ mice. genotypes (Supplementary Fig. 3G-H). mice. We then counted dendritic spine density only in double labeled We then hypothesized that, since Helios is expressed only in cal- (Dil / Calbindin) neurons from the CA1 pyramidal layer (Fig. 5A-B). We bindin-positive pyramidal CA1 neurons during hippocampal develop- observed that spine density in apical dendrites of CB-positive CA1 ment, it is conceivably that specific and localized architectural changes pyramidal neurons were decreased in He−/− mice compared to He+/+ could be found only in this neuronal subtype. Thus, we combined a mice (Fig. 5B-C). Further morphological analysis indicated that speci- DiOlistic labeling (Dil) with immunofluorescence against calbindin 1 in fically thin-like spines were the ones that were significantly reduced fixed hippocampal slices from adult (8-week-old) He+/+ and He−/− (Fig. 5D). As a control, we counted CB-positive CA1 pyramidal cell 10 A. Giralt, et al. Experimental Neurology 323 (2020) 113095 (caption on next page) 11 A. Giralt, et al. Experimental Neurology 323 (2020) 113095 Fig. 5. Characterization of structural and biochemical synaptic properties in He+/+ and He−/− mice. (A) Confocal microscopy images depicting the combination of DiI (reed) and immune-fluorescence for calbindin-D28k (green) in pyramidal neurons of the CA1 (8-weeks-old mice). Only pyramidal neurons co-labeled with DiI and calbindin-D28k (CB) were used for dendritic spine analysis. (B) Representative sections of 2nd-order apical dendrites from pyramidal CA1 CB-positive neurons in He+/+ and He−/− mice. (C) Quantitative analysis from B showing dendritic spine density per micrometer of dendritic length (t = 4.654, p < .001). (D) Percentage of each morphological type of dendritic spine (see Methods) from He+/+ and He−/− mice (genotype x spine type: F(2, 276) = 4.918, p < .01). (D) Quantification of CB-positive neuronal density in the pyramidal layer of the CA1 in 8-weeks-old He+/+ and He−/− mice (t = 1.199, p = .2647). (F) GluN1 (t = 4.061, p = .0036), GluN2A, GluN2B (t = 3.810, p = .0052) and GluN3A protein levels were analyzed in hippocampal samples from 8-week-old He+/+ and He−/− mice. Representative immunoblots are shown below the quantification graph for each protein. (G) PSD-95, Chapsyn110, SynGAP and CaMKIIα protein levels were analyzed in hippo- campal samples from 8-week-old He+/+ and He−/− mice. Representative immunoblots are shown below the quantification graph for each protein. (F) GluN1, GluN2A, GluN2B and GluN3A protein levels were analyzed in hippocampal samples from 8-days-old He+/+ and He−/− mice. Representative immunoblots are shown below the quantification graph for each protein. (G) PSD-95, Chapsyn110, SynGAP and CaMKIIα protein levels were analyzed in hippocampal samples from 8- days-old He+/+ and He−/− mice. Representative immunoblots are shown below the quantification graph for each protein. Bars represent mean ± SEM. In C and D He+/+ (n = 45 dendrites) and He−/− (n = 50 dendrites) dendrites were used to quantify spine density and density of spine types. In E n = 5 mice/genotype. In F-G n = 5 per genotype. In HeI, He+/+ n = 5 and He−/− n = 6. **: p < .01, ***: p < .001 with respect to He+/+ mice. Scale bar in A: 70 μm; Scale bar in B: 2 μm. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) density in both, He−/− compared to He+/+ mice. We did not find reach the maximum levels at postnatal day 15 (Fig. 6B). Looking at the significant changes between genotypes in this parameter (Fig. 5E). CA1 neurons in He+/+ pups at postnatal day 0 (P0) we found that We then sought for putative biochemical changes focusing on sy- VSNL1 was highly enriched in Helios-positive neurons of the hippo- naptic markers due to the alterations observed in LTP (Fig. 4D) and campal plate (Fig. 6C-D). We also observed that VSNL1 levels were spine density (Fig. 5C-D). Concretely, we evaluated the total protein significantly increased in neurons of the hippocampal plate in He−/− levels of different N-methyl-D-aspartate receptors subunits namely compared with He+/+ pups at P0 (Fig. 6C and E-F). GluN1, GluN2A, GluN2B and Glun3A (Fig. 5F) and postsynaptic density Thus, to evaluate whether developmental changes in VSNL1 levels markers namely PSD-95, Chapsyn110, CaMKIIα and SynGAP (Fig. 5G). could be relevant for the observed hippocampal-dependent phenotype We found that only GluN1 and GluN2B were significantly decreased in in He−/− mice we generated several plasmids to express shRNA to the hippocampus of adult He−/− compared with He+/+ mice (Fig. 5F). inhibit VSNL1 expression (Fig. 7A). From 5 different probes we found To gain insight whether these biochemical changes were persistent that the shRNA4 reduced approximately a 50% of VSNL1 protein ex- during the neonatal development or specific for adult ages, we analyzed pression (Fig. 7A-B) which could be appropriate to prevent the ~35% of the levels of all of them at postnatal day 8 (P8) in the hippocampus of VSNL1 increase observed in He−/− pups from Fig. 6. Next, we cultured He−/− He+/+ mice. Surprisingly we did not detect changes either in hippocampal primary neurons from He−/− and He+/+ embryos at E18 NMDARs subunits (Fig. 5H) or in postsynaptic density markers (Fig. 5I). which were co-transfected at day in vitro 21 (DIV21) with a GFP Altogether, these results suggest that the impact of the absence of He- plasmid (to label spines) and the scramble shRNA-mCherry or shRNA4- lios in hippocampal LTP and memory strongly correlates with a mod- mCherry to knock-down VSNL1 levels. After 72 h, we analyzed den- ulation of the synapse number and morphology specifically in CB-po- dritic spine density in selected neurons with pyramidal shape, co-ex- sitive pyramidal cells of the CA1. Furthermore, these changes strongly pressing GFP, mCherry and being positive for calbindin (Calbindin 1, correlate with a decrease on GluN1 and GluN2B protein levels in adult 28 kDa; CB-positive pyramidal neurons) (Fig. 7C-D). We found that, as but not neonatal hippocampi in Helios deficient mice indicating that observed in adult mice, He−/− CB-positive pyramidal neurons expres- these changes are not observable until the maturity is reached. sing scramble shRNA showed decreased spine density (Fig. 7E-F). This decrease was significantly restored in He−/− CB-positive pyramidal neurons expressing shRNA4 against VSNL1 (Fig. 7C-D). Altogether, 3.6. Identification of VSNL1 as a core gene to mediate hippocampal these results indicate that aberrant VSNL1 increased levels in He−/− maturation deficits in Helios deficient mice CB-positive pyramidal neurons are sufficient to induce a loss of its dendritic spine density. Our previous experiments indicate that He−/− mice develop func- tional, morphological and biochemical alterations dependent on the hippocampal function when they reach adulthood. However, Helios 4. Discussion expression is principally limited to developmental stages. Then, we wondered which type of molecular changes during Helios expression The heterogeneity of CA1-PNs of the hippocampus is increasingly could induce such long-term phenotypic outcomes. Then, we first per- recognized and theoretical and experimental studies suggest that formed a high throughput RNA sequencing in the hippocampus and learning and memory may be produced by the activity of discrete striatum of He−/− and He+/+ embryos (at embryonic day 18), when groups of neurons (Buzsaki, 2010; Tonegawa et al., 2015). In addition, Helios reaches its peak of expression. The results indicated that 6 hip- it is thought that some of these subpopulations may act as specific cells pocampal genes and 4 striatal genes significantly changed in He−/− to control the animal's ability to form spatial memory (Soltesz and compared with He+/+ E18 embryos including the Ikzf2 (Helios) gene Losonczy, 2018). However, the development of molecular, anatomical (Fig. 6A). From these genes, we found particularly interesting the in- and functional cues of these cells are still largely undescribed. Here we creased expression of the VSNL1 (Visinin-like 1 protein; VSNL1) gene show that a developmental affection of the transcription factor called since it is weakly/non-expressed in the striatum (Braunewell, 2012) and Helios in a subpopulation of calbindin positive-CA1-pyramidal neurons we found its up-regulation only in the hippocampus but not in the (CB-CA1-PNs) is critical for the correct acquisition of spatial memory, striatum of He−/− mice. Additionally, VSNL1 has been largely involved CA3-CA1 LTP expression and spine density and morphology. in synaptic plasticity and memory formation (Groblewska et al., 2015). Lamina-specific neurochemical differences have been described in This tissue-specificity could help to explain the divergence that we the CA1-PNs. Different CA1-PNs differ in molecular, structural and found between the striatal- and hippocampal-dependent phenotypes in physiological characteristics due to their differences in birthdate and He−/− mice. We then performed a time course only in He+/+ samples specific genetic programs (Soltesz and Losonczy, 2018). We show that of VSNL1 expression compared to Helios expression in the developing Helios is expressed in a specific subpopulation of CB-CA1-PNs. Our hippocampus and we found that they followed an opposite pattern. results demonstrate that CB-CA1-PNs actively participate in spatial Helios levels were high from embryos stages and then it progressively memory. We observe that aberrant development/maturation of Helios- decreased whereas, in contrast, VSNL1 levels progressively increased to expressing CB-CA1-PNs subpopulation produces dramatic affection of 12 A. Giralt, et al. Experimental Neurology 323 (2020) 113095 (caption on next page) 13 A. Giralt, et al. Experimental Neurology 323 (2020) 113095 Fig. 6. Identification and characterization of VSNL1 in the hippocampus of He+/+ and He−/− mice. (A) Embryonic (E18.5) hippocampal and striatal tissue from He+/+ and He−/− embryos were subjected to next-generation RNA sequencing. Blue values indicate down-regulated genes and reed values indicate up-regulated genes. 2-fold change and adjusted p value are depicted for each gene. (B) Time course of Helios and VSNL1 protein expression only in the He+/+ hippocampus. Upper panel shows representative immunoblots at embryonic day 18 and at postnatal days 0, 7, 15 and 40. Lower panel shows the corresponding densitometry quanti- fication. (C) Confocal microscopy images of CA1 immunostained for Helios (red) and VSNL1 (green) in He+/+ and He−/− pups at P0. Scale bar: 60 μm. (D) IOD quantification in He+/+ pups of VSNL1-positive labeling comparing Helios-positive cells versus Helios-negative cells (t = 17.38, p < .001). (E) IOD quantification of VSNL1-positive labeling from the MZ to the IZ of the CA1 of He+/+ and He−/− pups at P0. Using ROIs placed on the top of the dorsal CA1 (as depicted in C by dashed rectangles), we performed a linear intensity profile analysis as a mean IOD. Points in the X axis indicate mean IOD for each row of pixels. (F) Quantification of the HP mean IOD as in E (t = 3.242, p = .0088). See supplementary methods section for detailed statistics of the next-generation RNA sequencing in A. Bars represent mean ± SEM. In B n = 6 mice/group; in D n = 88 cells from 5 mice/genotype were evaluated; In E and F n = 6 mice/genotype. All data was analyzed using two- tailed Student's t-test. **: p < .01, ***: p < .001 with respect to He+/+ mice. IZ: Intermediate zone, dHP: Dorsal hippocampal plate, sHP: Superficial hippocampal plate. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) CA3-CA1 LTP. Accordingly, CA3 Schaffer collateral excitation is Varoqueaux et al., 2002; Verhage et al., 2000). In contrast, several stronger in CB-CA1-PNs (Valero et al., 2015) and spatial learning per- authors claim that initial development of dendrites, synapses and spines formances strongly correlate with CA3-CA1 LTP induction (Morris and are mostly controlled by cell-intrinsic genetic programs (Lu et al., Frey, 1997; Shapiro, 2001), suggesting a direct correlation between 2013). At the present time no specific intrinsic genetic program has affectation of LTP in Helios CB-CA1-PNs and spatial memory. been associated with spinogenic processes in CB-CA1-PNs. Here we We also found that Helios genetic deletion does not affect the gen- show that a specific transcription factor, Helios, is critical for the cor- eral dendritic arborization, but it induced dendritic spine number loss rect production of thin-like dendritic spines, suggesting an extremely specifically in CB-CA1-PNs. Supporting our results, it has been shown fine molecular regulation of the spinogenesis in the CB-CA1-PNs sub- that spatial memory and associative learning have been strongly asso- population. In this line, a putative cell intrinsic genetic program med- ciated with changes on hippocampal dendritic spines (Leuner et al., iating such phenomena could involve the regulation of intracellular 2003; Moser et al., 1994). Furthermore, specific dendritic spine loss in calcium homeostasis. We identified VSNL1, a neuronal calcium sensor CB-CA1-PNs could account for their essential role in the expression and protein, as the principal Helios downstream candidate that could ex- maintenance of hippocampal LTP and spatial learning (Jouvenceau plain our divergent striatal- and hippocampal-dependent phenotypes et al., 1999; Molinari et al., 1996). Similarly, the FMR1 KO mice, a since VSNL1 was observed to be increased only in the hippocampi from model of Fragile X syndrome (FXS), showed deficits in spatial learning He−/− mice compared to He+/+ mice. Indeed, VSNL1 expression is which were accompanied of altered functions in CB-CA1-PNs (Real almost absent in control striatal tissue (Paterlini et al., 2000). Within et al., 2011). On the other hand, present findings also demonstrate that the hippocampus, VSNL1 is expressed in pyramidal neurons of the CA1 the main synaptic morphological alteration was found in “thin-like” and in calretinin-positive interneurons (Bernstein et al., 2002). In spines whose are considered the learning spines due to their prominent contrast to Helios, VSNL1 expression increases progressively during motility (Bourne and Harris, 2007). Strengthening our data, it has been hippocampal development (present results) and it reaches its maximum shown that decreased density of thin spines is associated with impair- levels in adulthood (Gierke et al., 2004). We hypothesize that Helios ments in several cognitive tasks (Dumitriu et al., 2010) and treatments could modulate spinogenesis through a tight regulation of VSNL1 ex- aimed to counteract cognitive decline in neurological disorders result in pression. Reinforcing this idea, it has been proposed that during the an increase in numbers of thin spines (Bourne and Harris, 2007). maturation of neural circuits, the most prominent group of up-regulated Strikingly, we observed that early postnatal neurological functions genes consists on calcium-dependent membrane-binding proteins, in- were almost unaffected in He−/− mice whereas at young/adult stages, cluding VSNL1 (Yamatani et al., 2010). Although a direct Helios reg- He−/− mice displayed moderate and specific spatial learning and ulation of VSNL1 cannot be ruled out, Helios could also control VSNL1 memory impairments. These results are apparently opposed to the expression indirectly. Supporting the latter idea, it is known that Helios finding that hippocampal Helios expression peaks around peri-natal plays a role as a co-factor together with other transcription factors such stages, when CA1 neurons have finished their positioning and begin to as Nuclear Respiratory Factor-1 (NRF1) as demonstrated in previous mature (Altman and Bayer, 1990a; Altman and Bayer, 1990c; Altman analysis of DNA regions in which Helios showed significant enrichment and Bayer, 1990b; Bayer, 1980a; Bayer, 1980b; Danglot et al., 2006). of NRF1 binding motifs (Kim et al., 2015). NRF1 is a good candidate to However, we hypothesize that developmental deficiencies in Helios link Helios with VSNL1 since it is able to regulate VSNL1 gene tran- expressing CB-CA1-PNs induce long-term outcome effects that are only scription (Fu et al., 2009) and in turn, NRF1 is associated with Alz- observed when these hippocampal synapses are completely functional heimer's disease (Preciados et al., 2016) in correlation with the spatial at adult stages (Baudry et al., 1981; Dudek and Bear, 1993; Harris and learning deficits observed in mice devoid of Helios. We hypothesize Teyler, 1984) but not at early postnatal stages (Durand and Konnerth, then that a lack of the Helios repressive function on gene expression in 1996). In this line, CA1 synapses at first postnatal days are prominently He−/− mice would induce a premature increase on VSNL1 expression silent and it is hard to induce LTP on them during the first postnatal in CB-CA1-PNs which in turn would negatively influence the spine week, whereas robust LTP can be induced after the second week formation and maintenance. Similarly, overexpression of VSNL1 as well (Baudry et al., 1981; Harris and Teyler, 1984). Synaptogenesis occurs in as other calcium sensors are capable to reduce axon outgrowth and the CA1 from P2 onwards (Amaral and Dent, 1981; Buchs et al., 1993), branching in other neural systems (Angaut-Petit et al., 1998; Hui et al., and spines become more stable and more mature between P15 and P35 2007; Yamatani et al., 2010). VSNL1 overexpression might influence (Fiala et al., 1998; Harris et al., 1992). Supporting our hypothesis, the certain neuronal processes such as structural synaptic plasticity and dysregulation of other specific developmental expressed genes (Bhatt neuronal maturation via regulation of cAMP and cGMP production (Lin et al., 2009; van et al., 2013) or treatments with drugs during early life et al., 2002), (Braunewell et al., 2011a, 2011b). Accordingly, a sus- stages have been shown to induce long-term outcome effects in neural tained and occluded expression of the cAMP-PKA pathways can induce function and synaptic plasticity (Kataoka et al., 2013; Takuma et al., deficits in hippocampal-related learning and memory (Carlyle et al., 2014; Xiao et al., 2016). 2014; Giralt et al., 2011b; Ramos et al., 2003), possibly, via modulating The mechanism that instruct the assembly of fine-scale features of neuronal morphology during development (Cai et al., 2001). synaptic connectivity in CB-CA1-PNs are only beginning to be under- In summary, proper CB-CA1-PNs spine maturation depends on a stood. Growing evidences indicate that spines formation and synaptic homeostatic expression of VSNL1. VSNL1 levels during development connectivity occur in the absence of glutamate release (Lu et al., 2013; are tightly controlled by Helios transcription regulation in this neuronal 14 A. Giralt, et al. Experimental Neurology 323 (2020) 113095 (caption on next page) 15 A. Giralt, et al. Experimental Neurology 323 (2020) 113095 Fig. 7. Effects of VSNL1 levels normalization in He+/+ and He−/− CA1 CB-Positive Neurons. (A) Pyramidal neurons at day in vitro 21 were transfected with 5 different plasmids expressing shRNA constructs designed to prevent VSNL1 expression. (A and B) Note the significant reduction of VSNL1 expression provoked by the plasmid expressing the shRNA 4. The plasmid expressing the shRNA 4 was chosen for subsequent studies from B to F. Scale bar: 30 μm. (C) Pyramidal neurons at day in vitro 21 were co-transfected with a plasmid expressing GFP (green) and the plasmid expressing the shRNA 4 (reed/mCherry). 72 h after transfection cells were fixed and labeled for calbindin-D28k (CB). Scale bar: 50 μm. (C-D) Only cells with pyramidal-like shapes and co-labeled with GFP/mCherry/CB were used for the dendritic spine density analysis. Scale bar: 25 μm. (E) Representative dendrite sections labeled with GFP and used for dendritic spines density analysis. Scale bar: 3 μm. (F) Spine density quantification (Group effect: F(3,156) = 9.846, p < .001) of results as in E (n = 40 dendrites/group from 3 different cultures). Bars represent mean ± SEM. For statistical analysis one-way ANOVA with the Tukey's post hoc test was used. ***: p < .001 with respect to He+/+-Scr group; $$$: p < .001 with respect to He+/+-shRNA group; ###: p < .001 with respect to He−/−-Scr group. Scr: Scramble; shRNA: shRNA 4. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) subpopulation. An incorrect balance of this complex hippocampal López, Maria Teresa Muñoz for technical assistance. We also thank Dr. system during development would induce a long-lasting and specific Isabel Pérez Otaño (Centro de Investigaciones Médicas Aplicadas, incapacity to form spatial memory traces during adulthood as observed Pamplona. Spain) for her generous gift of the anti-GluN3A antibody. in several neuropsychiatric and cognitive disorders. The Rat-401 monoclonal anti-nestin antibody developed by S. Hockfield was obtained from the Development Studies Hybridoma Bank Ethics approval developed under the auspices of the NICHD and maintained by the Department of Biological Science, University of Iowa, Iowa City, Iowa. Mouse experiments were approved by local committees [Universitat de Barcelona, CEEA (133/10) and Generalitat de Catalunya (DAAM Appendix A. Supplementary data 5712)]. Supplementary data to this article can be found online at https:// Funding doi.org/10.1016/j.expneurol.2019.113095. This study was supported by grants from the Ministerio de References Economia y Competitividad (SAF 2014-57160-R to JA; BFU2017- 88393-P, to EDM; SAF2016-76340R to ES; and SAF2015-66505-R to Agoston, D.V., Szemes, M., Dobi, A., Palkovits, M., Georgopoulos, K., Gyorgy, A., Ring, J.M.C;), Spain; Instituto de Salud Carlos III, Ministerio de Economia y M.A., 2007. Ikaros is expressed in developing striatal neurons and involved in en- kephalinergic differentiation. J. Neurochem. 102, 1805–1816. Competitividad and European Regional Development Fund (ERDF) Altman, J., Bayer, S.A., 1990a. Migration and distribution of two populations of hippo- [CIBERNED, to JA and ES; Fondo de Investigaciones Sanitarias (PI11/ campal granule cell precursors during the perinatal and postnatal periods. J. Comp. 00704) to MP; and RETICS (RD16/0011/0011 to JCS; RD16/0011/ Neurol. 301, 365–381. Altman, J., Bayer, S.A., 1990b. Mosaic organization of the hippocampal neuroepithelium 0012 to JMC, Red de Terapia Celular), to JMC], Spain; Generalitat de and the multiple germinal sources of dentate granule cells. J. Comp. Neurol. 301, Catalunya (2017SGR-1095 to JA, and 2017SGR-1408 to JMC), Spain; 325–342. and ADVANCE(CAT) with the support of ACCIÓ (Catalonia Trade & Altman, J., Bayer, S.A., 1990c. Prolonged sojourn of developing pyramidal cells in the intermediate zone of the hippocampus and their settling in the stratum pyramidale. J. Investment; Generalitat de Catalunya), the European Community under Comp. Neurol. 301, 343–364. the Catalonian ERDF operational program 2014–2020], Spain and from Amaral, D.G., Dent, J.A., 1981. Development of the mossy fibers of the dentate gyrus: I. A Ministerio de Ciencia, Innovación y Universidades (RTI2018-094678-A- light and electron microscopic study of the mossy fibers and their expansions. J. Comp. Neurol. 195, 51–86. I00 to AG; RTI2018-099001-B-I00 to JMC). AG is a Ramón y Cajal Angaut-Petit, D., Toth, P., Rogero, O., Faille, L., Tejedor, F.J., Ferrus, A., 1998. Enhanced fellow (RYC-2016-19466). neurotransmitter release is associated with reduction of neuronal branching in a Drosophila mutant overexpressing frequenin. Eur. J. Neurosci. 10, 423–434. Baudry, M., Arst, D., Oliver, M., Lynch, G., 1981. Development of glutamate binding sites Authors' contributions and their regulation by calcium in rat hippocampus. Brain Res. 227, 37–48. Bayer, S.A., 1980a. Development of the hippocampal region in the rat. I. Neurogenesis A.G. conceived and carried out most experiments, analyzed and examined with 3H-thymidine autoradiography. J. Comp. Neurol. 190, 87–114. Bayer, S.A., 1980b. Development of the hippocampal region in the rat. II. Morphogenesis interpreted results and wrote the manuscript. V.B. carried out experi- during embryonic and early postnatal life. J. Comp. Neurol. 190, 115–134. ments related to gene gun and primary cultures. M.P. and R.-M.I. car- Bernstein, H.G., Braunewell, K.H., Spilker, C., Danos, P., Baumann, B., Funke, S., ried out the experiments in embryonic samples. S.R. carried out the Diekmann, S., Gundelfinger, E.D., Bogerts, B., 2002. Hippocampal expression of the Golgi Cox staining and some perinatal ICC experiments at embryonic calcium sensor protein visinin-like protein-1 in schizophrenia. Neuroreport 13, 393–396. stages and C.B. performed the electronic microscopy. L.M.-P. helped Bhatt, D.H., Zhang, S., Gan, W.B., 2009. Dendritic spine dynamics. Annu. Rev. Physiol. with the RNAseq experiments. A.Z.-M. And J.B. performed the elec- 71, 261–282. trophysiology experiments. E.B. and M.P. carried out the necropsy ex- Bourin, M., Hascoet, M., 2003. The mouse light/dark box test. Eur. J. Pharmacol. 463, 55–65. periments and tissue characterization of mutant mice. A.G.-T. And J.S. Bourne, J., Harris, K.M., 2007. Do thin spines learn to be mushroom spines that re- generated the VSNL1 plasmids. J.A. supervised the histological ex- member? Curr. Opin. Neurobiol. 17, 381–386. periments. S.G. supervised the biochemical experiments. M.P. su- Bourne, J.N., Harris, K.M., 2011. Coordination of size and number of excitatory and in- hibitory synapses results in a balanced structural plasticity along mature hippo- pervised the golgi experiments. E.S. supervised the electronic micro- campal CA1 dendrites during LTP. Hippocampus. 21 (4), 354–373. scopy experiments. E.D.M. supervised the electrophysiological Braunewell, K.H., 2012. The visinin-like proteins VILIP-1 and VILIP-3 in Alzheimer’s experiments. J.M.C. conceived and supervised the study, analyzed re- disease-old wine in new bottles. Front. Mol. Neurosci. 5, 20. Braunewell, K.H., Dwary, A.D., Richter, F., Trappe, K., Zhao, C., Giegling, I., Schönrath, sults and wrote the manuscript. K., Rujescu, D., 2011a. Association of VSNL1 with schizophrenia, frontal cortical function, and biological significance for its gene product as a modulator of cAMP Declaration of Competing Interest levels and neuronal morphology. Transl. Psychiatry 1, e22. Braunewell, K.H., Dwary, A.D., Richter, F., Trappe, K., Zhao, C., Giegling, I., Schonrath, K., Rujescu, D., 2011b. Association of VSNL1 with schizophrenia, frontal cortical The authors have no competing financial interest to declare. function, and biological significance for its gene product as a modulator of cAMP levels and neuronal morphology. Transl. Psychiatry 1, e22. Brito, V., Giralt, A., Enriquez-Barreto, L., Puigdellivol, M., Suelves, N., Zamora-Moratalla, Acknowledgements A., Ballesteros, J.J., Martin, E.D., Dominguez-Iturza, N., Morales, M., Alberch, J., Gines, S., 2014. Neurotrophin receptor p75(NTR) mediates Huntington’s disease-as- We are very grateful to Jean-Antoine Girault, Institut du Fer à sociated synaptic and memory dysfunction. J. Clin. Invest. 124, 4411–4428. Moulin (Paris, France) for his comments and advise. We thank Ana Brown, S.P., Hestrin, S., 2009. Cell-type identity: a key to unlocking the function of 16 A. Giralt, et al. Experimental Neurology 323 (2020) 113095 neocortical circuits. Curr. Opin. Neurobiol. 19, 415–421. Consortium, G.E.R.A.D., National Institute on Aging Late-Onset Alzheimer’s Disease Buchs, P.A., Stoppini, L., Muller, D., 1993. Structural modifications associated with sy- Family Study Group, Holmans, P., Jones, L., Devlin, B., Klei, L., Barmada, M.M., naptic development in area CA1 of rat hippocampal organotypic cultures. Brain Res. Demirci, F.Y., De Kosky, S.T., Lopez, O.L., Passmore, P., Owen, M.J., O’Donovan, Dev. Brain Res. 71, 81–91. M.C., Mayeux, R., Kamboh, M.I., Williams, J., 2012. Genome-wide association study Buzsaki, G., 2010. Neural syntax: cell assemblies, synapsembles, and readers. Neuron 68, of Alzheimer’s disease with psychotic symptoms. Mol. Psychiatry 1316–1327 (12). 362–385. Huang, C., Chen, M., Pang, D., Bi, D., Zou, Y., Xia, X., Yang, W., Luo, L., Deng, R., Tan, H., Cai, D., Qiu, J., Cao, Z., McAtee, M., Bregman, B.S., Filbin, M.T., 2001. Neuronal cyclic Zhou, L., Yu, S., Guo, L., Du, X., Cui, Y., Hu, J., Mao, Q., Worley, P.F., Xiao, B., 2014. AMP controls the developmental loss in ability of axons to regenerate. J. Neurosci. Developmental and activity-dependent expression of LanCL1 confers antioxidant 21, 4731–4739. activity required for neuronal survival. Dev. Cell 30, 479–487. Cai, Q., Dierich, A., Oulad-Abdelghani, M., Chan, S., Kastner, P., 2009. Helios deficiency Hui, K., Fei, G.H., Saab, B.J., Su, J., Roder, J.C., Feng, Z.P., 2007. Neuronal calcium has minimal impact on T cell development and function. J. Immunol. 183, sensor-1 modulation of optimal calcium level for neurite outgrowth. Development 2303–2311. 134, 4479–4489. Carlyle, B.C., Nairn, A.C., Wang, M., Yang, Y., Jin, L.E., Simen, A.A., Ramos, B.P., Jammalamadaka, A., Banerjee, S., Manjunath, B.S., Kosik, K.S., 2013. Statistical analysis Bordner, K.A., Craft, G.E., Davies, P., Pletikos, M., Sestan, N., Arnsten, A.F., Paspalas, of dendritic spine distributions in rat hippocampal cultures. BMC. Bioinforma. 14, C.D., 2014. cAMP-PKA phosphorylation of tau confers risk for degeneration in aging 287. association cortex. Proc. Natl. Acad. Sci. U. S. A. 111, 5036–5041. Jouvenceau, A., Potier, B., Battini, R., Ferrari, S., Dutar, P., Billard, J.M., 1999. Cembrowski, M.S., Bachman, J.L., Wang, L., Sugino, K., Shields, B.C., Spruston, N., 2016. Glutamatergic synaptic responses and long-term potentiation are impaired in the CA1 Spatial gene-expression gradients underlie prominent heterogeneity of CA1 pyr- hippocampal area of calbindin D(28k)-deficient mice. Synapse 33, 172–180. amidal neurons. Neuron 89, 351–368. Kataoka, S., Takuma, K., Hara, Y., Maeda, Y., Ago, Y., Matsuda, T., 2013. Autism-like Cobb, B.S., Smale, S.T., 2005. Ikaros-family proteins: in search of molecular functions behaviours with transient histone hyperacetylation in mice treated prenatally with during lymphocyte development. Curr. Top. Microbiol. Immunol. 290, 29–47. valproic acid. Int. J. Neuropsychopharmacol. 16, 91–103. Danglot, L., Triller, A., Marty, S., 2006. The development of hippocampal interneurons in Kim, H.J., Barnitz, R.A., Kreslavsky, T., Brown, F.D., Moffett, H., Lemieux, M.E., rodents. Hippocampus 16, 1032–1060. Kaygusuz, Y., Meissner, T., Holderried, T.A., Chan, S., Kastner, P., Haining, W.N., Deguchi, Y., Donato, F., Galimberti, I., Cabuy, E., Caroni, P., 2011. Temporally matched Cantor, H., 2015. Stable inhibitory activity of regulatory T cells requires the tran- subpopulations of selectively interconnected principal neurons in the hippocampus. scription factor Helios. Science 350, 334–339. Nat. Neurosci. 14, 495–504. Leuner, B., Falduto, J., Shors, T.J., 2003. Associative memory formation increases the Dierssen, M., Fotaki, V., Martinez de, L.M., Gratacos, M., Arbones, M., Fillat, C., Estivill, observation of dendritic spines in the hippocampus. J. Neurosci. 23, 659–665. X., 2002. Neurobehavioral development of two mouse lines commonly used in Lex, B., Sommer, S., Hauber, W., 2011. The role of dopamine in the dorsomedial striatum transgenic studies. Pharmacol. Biochem. Behav. 73, 19–25. in place and response learning. Neuroscience 172, 212–218. Dimsdale-Zucker, H.R., Ritchey, M., Ekstrom, A.D., Yonelinas, A.P., Ranganath, C., 2018. Li, J.T., Zhao, Y.Y., Wang, H.L., Wang, X.D., Su, Y.A., Si, T.M., 2015. Long-term effects of CA1 and CA3 differentially support spontaneous retrieval of episodic contexts within neonatal exposure to MK-801 on recognition memory and excitatory-inhibitory bal- human hippocampal subfields. Nat. Commun. 9 (1), 294 18. ance in rat hippocampus. Neuroscience 308, 134–143. Dong, H.W., Swanson, L.W., Chen, L., Fanselow, M.S., Toga, A.W., 2009. Genomic-ana- Lin, L., Braunewell, K.H., Gundelfinger, E.D., Anand, R., 2002. Functional analysis of tomic evidence for distinct functional domains in hippocampal field CA1. Proc. Natl. calcium-binding EF-hand motifs of visinin-like protein-1. Biochem. Biophys. Res. Acad. Sci. U. S. A. 106, 11794–11799. Commun. 296, 827–832. Dudek, S.M., Bear, M.F., 1993. Bidirectional long-term modification of synaptic effec- Lu, W., Bushong, E.A., Shih, T.P., Ellisman, M.H., Nicoll, R.A., 2013. The cell-autonomous tiveness in the adult and immature hippocampus. J. Neurosci. 13, 2910–2918. role of excitatory synaptic transmission in the regulation of neuronal structure and Dumitriu, D., Hao, J., Hara, Y., Kaufmann, J., Janssen, W.G., Lou, W., Rapp, P.R., function. Neuron 78, 433–439. Morrison, J.H., 2010. Selective changes in thin spine density and morphology in Martin-Ibanez, R., Crespo, E., Urban, N., Sergent-Tanguy, S., Herranz, C., Jaumot, M., monkey prefrontal cortex correlate with aging-related cognitive impairment. J. Valiente, M., Long, J.E., Pineda, J.R., Andreu, C., Rubenstein, J.L., Marin, O., Neurosci. 30, 7507–7515. Georgopoulos, K., Mengod, G., Farinas, I., Bachs, O., Alberch, J., Canals, J.M., 2010. Durand, G.M., Konnerth, A., 1996. Long-term potentiation as a mechanism of functional Ikaros-1 couples cell cycle arrest of late striatal precursors with neurogenesis of en- synapse induction in the developing hippocampus. J. Physiol. Paris 90, 313–315. kephalinergic neurons. J. Comp. Neurol. 518, 329–351. Fiala, J.C., Feinberg, M., Popov, V., Harris, K.M., 1998. Synaptogenesis via dendritic fi- Martin-Ibanez, R., Crespo, E., Esgleas, M., Urban, N., Wang, B., Waclaw, R., lopodia in developing hippocampal area CA1. J. Neurosci. 18, 8900–8911. Georgopoulos, K., Martinez, S., Campbell, K., Vicario-Abejon, C., Alberch, J., Chan, Fiala, J.C., Spacek, J., Harris, K.M., 2002. Dendritic spine pathology: cause or con- S., Kastner, P., Rubenstein, J.L., Canals, J.M., 2012. Helios transcription factor ex- sequence of neurological disorders? Brain Res. Brain Res. Rev. 39, 29–54. pression depends on Gsx2 and Dlx1&2 function in developing striatal matrix neurons. Fox, W.M., 1965. Reflex-ontogeny and behavioural development of the mouse. Anim. Stem Cells Dev. 21, 2239–2251. Behav. 13, 234–241. Metz, G.A., Schwab, M.E., 2004. Behavioral characterization in a comprehensive mouse Fu, J., Zhang, J., Jin, F., Patchefsky, J., Braunewell, K.H., Klein-Szanto, A.J., 2009. test battery reveals motor and sensory impairments in growth-associated protein-43 Promoter regulation of the visinin-like subfamily of neuronal calcium sensor proteins null mutant mice. Neuroscience 129, 563–574. by nuclear respiratory factor-1. J. Biol. Chem. 284, 27577–27586. Molinari, S., Battini, R., Ferrari, S., Pozzi, L., Killcross, A.S., Robbins, T.W., Jouvenceau, Gierke, P., Zhao, C., Brackmann, M., Linke, B., Heinemann, U., Braunewell, K.H., 2004. A., Billard, J.M., Dutar, P., Lamour, Y., Baker, W.A., Cox, H., Emson, P.C., 1996. Expression analysis of members of the neuronal calcium sensor protein family: Deficits in memory and hippocampal long-term potentiation in mice with reduced combining bioinformatics and Western blot analysis. Biochem. Biophys. Res. calbindin D28K expression. Proc. Natl. Acad. Sci. U. S. A. 93, 8028–8033. Commun. 323, 38–43. Morris, R.G., Frey, U., 1997. Hippocampal synaptic plasticity: role in spatial learning or Giralt, A., Friedman, H.C., Caneda-Ferron, B., Urban, N., Moreno, E., Rubio, N., Blanco, J., the automatic recording of attended experience? Philos. Trans. R. Soc. Lond. Ser. B Peterson, A., Canals, J.M., Alberch, J., 2010. BDNF regulation under GFAP promoter Biol. Sci. 352, 1489–1503. provides engineered astrocytes as a new approach for long-term protection in Moser, M.B., Trommald, M., Andersen, P., 1994. An increase in dendritic spine density on Huntington's disease. Gene Ther. 17, 1294–1308. hippocampal CA1 pyramidal cells following spatial learning in adult rats suggests the Giralt, A., Carreton, O., Lao-Peregrin, C., Martin, E.D., Alberch, J., 2011a. Conditional formation of new synapses. Proc. Natl. Acad. Sci. U. S. A. 91, 12673–12675. BDNF release under pathological conditions improves Huntington's disease pathology Packard, M.G., Knowlton, B.J., 2002. Learning and memory functions of the basal ganglia. by delaying neuronal dysfunction. Mol. Neurodegener. 6, 71. Annu. Rev. Neurosci. 25, 563–593. Giralt, A., Saavedra, A., Carreton, O., Xifro, X., Alberch, J., Perez-Navarro, E., 2011b. Paterlini, M., Revilla, V., Grant, A.L., Wisden, W., 2000. Expression of the neuronal cal- Increased PKA signaling disrupts recognition memory and spatial memory: role in cium sensor protein family in the rat brain. Neuroscience 99, 205–216. Huntington's disease. Hum. Mol. Genet. 20, 4232–4247. Preciados, M., Yoo, C., Roy, D., 2016. Estrogenic endocrine disrupting chemicals influ- Giralt, A., Sanchis, D., Cherubini, M., Ginés, S., Cañas, X., Comella, J.X., Alberch, J., 2013. encing NRF1 regulated gene networks in the development of complex human brain Neurobehavioral characterization of endonuclease G knockout mice reveals a new diseases. Int. J. Mol. Sci. 17. putative molecular player in the regulation of anxiety. Exp. Neurol. 247, 122–129. Ramos, B.P., Birnbaum, S.G., Lindenmayer, I., Newton, S.S., Duman, R.S., Arnsten, A.F., Giralt, A., Brito, V., Chevy, Q., Simonnet, C., Otsu, Y., Cifuentes-Diaz, C., de, P.B., Coura, 2003. Dysregulation of protein kinase a signaling in the aged prefrontal cortex: new R., Alberch, J., Gines, S., Poncer, J.C., Girault, J.A., 2017. Pyk2 modulates hippo- strategy for treating age-related cognitive decline. Neuron 40, 835–845. campal excitatory synapses and contributes to cognitive deficits in a Huntington's Real, M.A., Simon, M.P., Heredia, R., de, D.Y., Guirado, S., 2011. Phenotypic changes in disease model. Nat. Commun. 8, 15592. calbindin D28K immunoreactivity in the hippocampus of Fmr1 knockout mice. J. Groblewska, M., Muszynski, P., Wojtulewska-Supron, A., Kulczynska-Przybik, A., Comp. Neurol. 519, 2622–2636. Mroczko, B., 2015. The role of Visinin-like Protein-1 in the pathophysiology of Sellami, A., Al Abed, A.S., Brayda-Bruno, L., Etchamendy, N., Valério, S., Oulé, M., Alzheimer’s disease. J. Alzheimers Dis. 47, 17–32. Pantaléon, L., Lamothe, V., Potier, M., Bernard, K., Jabourian, M., Herry, C., Mons, Harris, K.M., Teyler, T.J., 1984. Developmental onset of long-term potentiation in area N., Piazza, P.V., Eichenbaum, H., Marighetto, A., 2017. Temporal binding function of CA1 of the rat hippocampus. J. Physiol. 346, 27–48. dorsal CA1 is critical for declarative memory formation. Proc. Natl. Acad. Sci. U. S. A. Harris, K.M., Jensen, F.E., Tsao, B., 1992. Three-dimensional structure of dendritic spines 114 (38), 10262–10267. and synapses in rat hippocampus (CA1) at postnatal day 15 and adult ages: im- Shapiro, M., 2001. Plasticity, hippocampal place cells, and cognitive maps. Arch. Neurol. plications for the maturation of synaptic physiology and long-term potentiation. J. 58, 874–881. Neurosci. 12, 2685–2705. Slomianka, L., Amrein, I., Knuesel, I., Sorensen, J.C., Wolfer, D.P., 2011. Hippocampal Hollingworth, P., Sweet, R., Sims, R., Harold, D., Russo, G., Abraham, R., Stretton, A., pyramidal cells: the reemergence of cortical lamination. Brain Struct. Funct. 216, Jones, N., Gerrish, A., Chapman, J., Ivanov, D., Moskvina, V., Lovestone, S., Priotsi, 301–317. P., Lupton, M., Brayne, C., Gill, M., Lawlor, B., Lynch, A., Craig, D., McGuinness, B., Soltesz, I., Losonczy, A., 2018. CA1 pyramidal cell diversity enabling parallel information Johnston, J., Holmes, C., Livingston, G., Bass, N.J., Gurling, H., McQuillin, A., processing in the hippocampus. Nat. Neurosci. 21, 484–493. 17 A. Giralt, et al. Experimental Neurology 323 (2020) 113095 Stevenson, R.F., Zheng, J., Mnatsakanyan, L., Vadera, S., Knight, R.T., Lin, J.J., Yassa, van, L.H., de Haan, T.R., Offringa, M., Post, B., van der Lee, J.H., 2013. Prognostic tests in M.A., 2018. Hippocampal CA1 gamma power predicts the precision of spatial term neonates with hypoxic-ischemic encephalopathy: a systematic review. memory judgments. Proc. Natl. Acad. Sci. U. S. A. 115 (40), 10148–10153. Pediatrics 131, 88–98. Takuma, K., Hara, Y., Kataoka, S., Kawanai, T., Maeda, Y., Watanabe, R., Takano, E., Varoqueaux, F., Sigler, A., Rhee, J.S., Brose, N., Enk, C., Reim, K., Rosenmund, C., 2002. Hayata-Takano, A., Hashimoto, H., Ago, Y., Matsuda, T., 2014. Chronic treatment Total arrest of spontaneous and evoked synaptic transmission but normal synapto- with valproic acid or sodium butyrate attenuates novel object recognition deficits and genesis in the absence of Munc13-mediated vesicle priming. Proc. Natl. Acad. Sci. U. hippocampal dendritic spine loss in a mouse model of autism. Pharmacol. Biochem. S. A. 99, 9037–9042. Behav. 126, 43–49. Venerosi, A., Valanzano, A., Alleva, E., Calamandrei, G., 2001. Prenatal exposure to anti- Thompson, C.L., Pathak, S.D., Jeromin, A., Ng, L.L., MacPherson, C.R., Mortrud, M.T., HIV drugs: neurobehavioral effects of zidovudine (AZT) + lamivudine (3TC) treat- Cusick, A., Riley, Z.L., Sunkin, S.M., Bernard, A., Puchalski, R.B., Gage, F.H., Jones, ment in mice. Teratology 63, 26–37. A.R., Bajic, V.B., Hawrylycz, M.J., Lein, E.S., 2008. Genomic anatomy of the hippo- Verhage, M., Maia, A.S., Plomp, J.J., Brussaard, A.B., Heeroma, J.H., Vermeer, H., campus. Neuron 60, 1010–1021. Toonen, R.F., Hammer, R.E., van den Berg, T.K., Missler, M., Geuze, H.J., Sudhof, Tonegawa, S., Liu, X., Ramirez, S., Redondo, R., 2015. Memory engram cells have come of T.C., 2000. Synaptic assembly of the brain in the absence of neurotransmitter se- age. Neuron 87, 918–931. cretion. Science 287, 864–869. Valero, M., Cid, E., Averkin, R.G., Aguilar, J., Sanchez-Aguilera, A., Viney, T.J., Gomez- Xiao, H., Liu, B., Chen, Y., Zhang, J., 2016. Learning, memory and synaptic plasticity in Dominguez, D., Bellistri, E., de la Prida, L.M., 2015. Determinants of different deep hippocampus in rats exposed to sevoflurane. Int. J. Dev. Neurosci. 48, 38–49. and superficial CA1 pyramidal cell dynamics during sharp-wave ripples. Nat. Yamatani, H., Kawasaki, T., Mita, S., Inagaki, N., Hirata, T., 2010. Proteomics analysis of Neurosci. 18, 1281–1290. the temporal changes in axonal proteins during maturation. Dev. Neurobiol. 70, Van Raamsdonk, J.M., Pearson, J., Slow, E.J., Hossain, S.M., Leavitt, B.R., Hayden, M.R., 523–537. 2005. Cognitive dysfunction precedes neuropathology and motor abnormalities in Yoshimura, Y., Callaway, E.M., 2005. Fine-scale specificity of cortical networks depends the YAC128 mouse model of Huntington's disease. J. Neurosci. 25, 4169–4180. on inhibitory cell type and connectivity. Nat. Neurosci. 8, 1552–1559. 18