Microglial Caspase-3 is essential for modulating hippocampal neurogenesis.

Brain, Behavior, and Immunity 112 (2023) 206–219 Available online 15 June 2023 0889-1591/© 2023 The Author(s). Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/bync-nd/4.0/). Microglial Caspase-3 is essential for modulating hippocampal neurogenesis Isabel M. Alonso Bellido a , b , Mercedes Posada-P´ erez a , c , Francisco Hern´ andez-Rasco a , b , Sandra V´ azquez-Reyes a , b , María Cabanillas a , b , Antonio J. Herrera a , b , Sara Bachiller a , d , e , Jesús Sold´ an-Hidalgo a , b , Ana M. Espinosa-Oliva a , b , Bertrand Joseph c , Rocío M. de Pablos a , b , Jos´ e L. Venero a , b , 1 , Rocío Ruiz a , b , 1 , * a Instituto de Biomedicina de Sevilla, IBiS/Hospital Universitario Virgen del Rocío/CSIC/Universidad de Sevilla, Sevilla, Spain b Departamento de Bioquímica y Biología Molecular, Facultad de Farmacia, Universidad de Sevilla, Spain c Institute of Environmental Medicine, Toxicology Unit, Karolinska Institutet, Stockholm, Sweden d Clinical Unit of Infectious Diseases, Microbiology and Parasitology, Laboratory of Immunovirology, Virgen del Rocío University Hospital, Seville, Spain e Department of Medical Biochemistry, Molecular Biology and Immunology, School of Medicine, University of Seville, Seville, Spain ARTICLE INFO Keywords: Caspase Neurogenesis Microglia Cognitive impairment Behaviour Hippocampus ABSTRACT Adult hippocampal neurogenesis (AHN) is a process involved in numerous neurodegenerative diseases. Many researchers have described microglia as a key component in regulating the formation and migration of new neurons along the rostral migratory stream. Caspase-3 is a cysteine-aspartate-protease classically considered as one of the main effector caspases in the cell death program process. In addition to this classical function, we have identified the role of this protein as a modulator of microglial function; however, its action on neurogenic processes is unknown. The aim of the present study is to identify the role of Caspase-3 in neurogenesis-related microglial functions. To address this study, Caspase-3 conditional knockout mice in the microglia cell line were used. Using this tool, we wanted to elucidate the role of this protein in microglial function in the hippocampus, the main region in which adult neurogenesis takes place. After the reduction of Caspase-3 in microglia, mutant mice showed a reduction of microglia in the hippocampus, especially in the dentate gyrus region, a region inherently associated to neurogenesis. In addition, we found a reduction in doublecortin-positive neurons in conditional Caspase-3 knockout mice, which corresponds to a reduction in neurogenic neurons. Furthermore, using high-resolution image analysis, we also observed a reduction in the phagocytic capacity of microglia lacking Caspase-3. Behavioral analysis using object recognition and Y-maze tests showed altered memory and learning in the absence of Caspase-3. Finally, we identified specific microglia located specifically in neurogenic niche positive for Galectin 3 which colocalized with Cleaved-Caspase-3 in control mice. Taken together, these results showed the essential role of Caspase-3 in microglial function and highlight the relevant role of this specific microglial phenotype in the maintenance of AHN in the hippocampus. 1. Introduction Neurogenesis, defined as the process of generation of new functional neurons from precursors, is restricted in the adult brain under baseline conditions in two neurogenic zones initially defined by Altman and Das: the subventricular zone (SVZ) of the lateral ventricle and the subgranular zone (SGZ) of the hippocampus (Altman and Das, 1965). Adult hippocampal neurogenesis (AHN) is a phenomenon that occurs in both, rodents (Ehninger and Kempermann, 2008) and humans (Cipriani et al., 2018; Moreno-Jim´ enez et al., 2019). Currently, newly generated neurons are believed to contribute to hippocampal-dependent learning and memory, among other functions (Anacker and Hen, 2017; Miller and Sahay, 2019). Since the existence of these neurogenic niches was first described, many researchers have attempted to study the modulation of this process. Consequently, neurogenesis alterations have been described in chronic neurodegenerative diseases (for a review, see: (Horgusluoglu et al., 2017)) such as Alzheimer’s disease (Cipriani et al., 2018; Moreno-Jim´ enez et al., 2019), Creutzfeldt-Jakob prion disease (Rela˜ no-Gin´ es et al., 2014) and Parkinson’s disease (H¨ oglinger et al., 2004; Kohl et al., 2012) and in acute diseases such as acute cerebral * Corresponding author. E-mail address: [email protected] (R. Ruiz). 1 Shared senior authorship. Contents lists available at ScienceDirect Brain Behavior and Immunity journal homepage: www.elsevier.com/locate/ybrbi https://doi.org/10.1016/j.bbi.2023.06.013 Received 3 January 2023; Received in revised form 7 June 2023; Accepted 10 June 2023 Brain Behavior and Immunity 112 (2023) 206–219 207 stroke (Lindvall and Kokaia, 2015). The proliferation, survival, differentiation and integration of new neurons in the adult hippocampus are processes regulated by multiple factors. In this sense, microglia, the immune cells of the nervous system, are an essential cellular factor for the generation and integration of new neurons (P´ erez-Rodríguez et al., 2021). Recently, a new type of proliferative region-associated microglia (PAM) has been described, which is especially evident during early postnatal development. This phenotype has a specific genetic signature that shares many molecular markers with disease-associated microglia (DAM), such as Gpnmb, Spp1, Igf1, Clec7a and Lgals3, among others (Li et al., 2019; Keren-Shaul et al., 2017; Krasemann et al., 2017) and is located in white matter and in neurogenic niches, suggesting the existence of a specific molecular machinery in microglial cells that modulates the proliferation of the newly synthesized neurons. Thus, this cell population releases several factors that modulate the proliferation or survival of different cells. In fact, some in vitro studies suggest that cultured microglia promotes precursor cell differentiation (Aarum et al., 2003) and increases neuroblast production and survival of adult neurons (Walton et al., 2006). However, the role of microglia in this process can be ambivalent. These cells have also been described to play a detrimental role in neurogenesis under pathological conditions through the excessive and chronic release of inflammatory cytokines such as interleukin (IL)-1β, tumor necrosis factor alpha (TNF α ) and IL-6 (Monje and Palmer, 2003; Musaelyan et al., 2014). Microglia are also responsible for the elimination of excess new-born cells through phagocytosis to maintain homeostasis (Sierra et al., 2010). Since the mechanisms underlying the regulation of hippocampal neurogenesis by microglia have not yet been explored sufficiently (Sierra et al., 2014), it is necessary to design new experiments to clarify the role of modulators generated by microglia, as well as the regulation of its production and release. Caspases are a group of intracellular cysteine proteases that play a crucial role in various biological processes, such as apoptosis (programmed cell death), necrosis, proliferation, differentiation, and inflammation. They are initially synthesized as inactive procaspases, which can be activated by different stimuli. Among caspases, the apoptotic caspases can be categorized into two distinct classes: the initiator or apical apoptotic caspases, including caspases-2, −8, −9, and −10, and the effector or executioner apoptotic caspases, including caspases-3 (Casp3), −6, and −7 (Shen et al., 2018; Venero et al., 2013). Among these, Casp3 is the most extensively studied and the most important executioner apoptotic caspase. Caspases play a crucial role in apoptotic cell death, but recent research suggests that they also regulate non-apoptotic functions in immune cells, including microglia. Our research has uncovered a previously unknown role for caspases-8, −3, and −7 in the pro-inflammatory activation of microglia (Burguillos et al., 2011). Surprisingly, we found that apoptotic caspases can regulate microglia to adopt a pro-inflammatory phenotype, and this effect was linked to a modest increase in Casp3 activity (Burguillos et al., 2011). Further investigations from our team revealed that inhibiting basal Casp3 activity in microglia resulted in the polarization of microglia towards a tumor-supportive phenotype (Shen et al., 2016a). These findings indicate that Casp3 functions as a modulator that regulates the response of microglial subtypes to different stimuli. In summary, while caspases are typically associated with apoptotic cell death, our research highlights their important regulatory role in immune cells, particularly in microglia, where Casp3 activity appears to be a key modulator of microglial subtypes in response to various stimuli. An association has been found between a specific subtype of microglia and neurogenic niches, indicating its key role in modulating adult neurogenesis beyond the removal of apoptotic cells (as noted in (Sierra et al., 2014)). This modulation is achieved through the production of factors that induce the proliferation and differentiation of precursor cells (Aarum et al., 2003) as well as the production of neuroblasts (Walton et al., 2006). Furthermore, it has been demonstrated that chronic disruption of microglial phagocytosis can impair neurogenesis (as shown in (Diaz-Aparicio et al., 2020)). Overall, these findings highlight the importance of microglia in regulating adult neurogenesis. Consequently, we investigated whether Casp3 activity plays a significant role in modulating microglia associated with neurogenic niches, and by extension, if it exerts an effect on adult neurogenesis. Consequently, the aim of the present study was to identify the role of Casp3 in neurogenesis-related microglial function. To this end, we took advantage of conditional knockout mice for Casp3 using the microgliaspecific Cx3cr1 Cre line (Casp3-cKO) in order to elucidate the role of this protein in microglial function in the hippocampus, the main neurogenic region of the brain. Following this approach, we found that Casp3-cKO mice showed a reduction in AHN. This reduction in AHN was inversely correlated with the density of microglia in the hippocampus. Importantly, we identified a microglia subtype that is linked to the hippocampal neurogenic niche in adult animals. This subtype is characterized by the expression of galectin-3 (Gal3), which is encoded by the Lgals3 gene. Remarkably, the absence of Casp3 in microglia led to a reduction in the number of microglia expressing Gal3 in the hippocampus, and, most importantly, a decrease in neurogenesis along with alterations in learning and memory. 2. Material and methods 2.1. Animal model We have used a conditional animal model that allows Casp3 deletion only in microglial cells. Caspase-3 f/f C57BL/6 mice with the Casp3 allele floxed at exon 2 were generously provided by Prof. Richard Flavell (Yale University). C57BL/6 mice containing a CRE recombinase under the control of the microglial Cx3cr1 promoter were acquired at The Jackson Laboratory (#21160, USA). These mice lines were intercrossed until the experimental Cx3cr1 Cre/WT -Casp3 f/f (Casp3-cKO) and control Cx3cr1 WT/ WT -Casp3 f/f (Casp3-Control) mice were obtained. This CRE system is inducible by tamoxifen, although we have also found some YFP expression in the absence of tamoxifen (Supplementary Fig. S3). Therefore, 3–4-month-old mice were injected intraperitoneally with tamoxifen (5% corn oil-ethanol at 2 mg/ml) daily for 5 days. All experiments were carried out three weeks after injection of tamoxifen (Fig. 2A). Both colonies were maintained at the Center for Animal Production and Experimentation of the University of Seville. The animals were housed at a constant room temperature of 22 ±1 ◦C and relative humidity (60%), with a 12 h light–dark cycle and ad libitum access to food and water. The experiments were carried out according to the Guidelines of the European Union Directive (2010/63/EU) and the Spanish regulations (BOE 34/11370–421, 2013) for the use of laboratory animals; the study was approved by the Scientific Committee of the University of Seville. 2.2. Behavioral tests Behavioral tests were videotaped in 4- to 5-month-old male mice, and subsequently, both in the open field test and in the Y-maze and object recognition tests, a series of parameters related to mouse movement were analyzed with Smart® software (SMART VIDEO TRACKING Software-Panlab | Harvard Apparatus). Behavioral experiments were conducted in behavioral test rooms during the standard light phase. Mice were brought to a test room at least 30 min before the behavioral test was performed. All work equipment was thoroughly cleaned with 70% ethanol between tests. Each animal performed the different behavioral tests in ascending order of stress testing (García-Domínguez et al., 2021; Su´ arez-Pereira et al., 2022). 2.3. Open field Mice were placed in the 55 ×40 ×40 cm arena for 5 min. This test provides a measure of anxiety as well as a measure of the locomotor I.M. Alonso Bellido et al. Brain Behavior and Immunity 112 (2023) 206–219 208 activity of the animals. Specifically, the time the mouse spent on the border of the arena, the resting time, and the number of times it explored the central region were measured. These measures were automatic through the software indicated above. 2.4. Rotarod The rotarod (Ugo Basile Biological Research Apparatus, France) consists of a horizontal cylinder 3 cm in diameter and 30 cm in length that can rotate at fixed or accelerated speed in a controlled manner. Motor coordination and balance were evaluated on a rotarod apparatus (Ugo Basile). Mice were placed on the rotarod at a constant speed (5 rpm) for a maximum of 5 min and at an accelerated speed (4 to 40 rpm in 5 min). Mice were allowed to perform 4 trials with 15 min of rest between trials during the first day. On the second day, a final test was performed. The number of falls and the latency to fall were recorded. 2.5. Marble burying For this test, mice were placed in a 20 ×36 ×14 cm arena with 2 cm of sawdust and 12 marbles arranged in 3 columns and 4 rows. The mice were placed in the arena and allowed to explore freely for 10 min. Data were taken on the number of marbles buried, considering them as those with<25% of the surface area uncovered. 2.6. Y-Maze This test uses a maze with three equal arms (8 ×18 ×38 cm). The mice were placed in the center of the maze at the beginning of the test, first allowing free exploration for 5 min. One hour later, a door was placed blocking one of the arms and they were allowed to explore for 5 min. After one hour, the door that blocked the arm was removed to allow free exploration again for another 5 min. In this test, the spontaneous alternation triplet and the number of arm entries were measured. 2.7. Object recognition memory (ORM) Mice were placed in the arena (38 ×15x 25 cm) for 5 min as a mode of adaptation to the environment. Then, two identical objects (A-A) were placed and recorded and left to explore freely for 15 min. One hour after training, the mice were placed in the arena with a familiar object (constant object-A) and an object different from the previous ones (novel object-B) for 10 min. Relative exploration of the novel object was expressed as a discrimination index (#novel - #familiar)/(#novel + #familiar), taking into account the training index. 2.8. Sacrifice and tissue extraction After behavioral tests, mice were anesthetized with 0.25 ml of ketamine (domtor 1 mg/ml, ketamidor 100 mg/ml, nosedorm 5 mg/ml, and bupaq 0.3 mg/ml). This was followed by intracardiac perfusion with 0.9% saline. Once perfusion was complete, the mice were decapitated and their brains were extracted. Each of the hemispheres was used for different purposes. The right hemisphere was fixed with 4% paraformaldehyde for four days at 4 ◦C and then cryoprotected in a 30% sucrose solution at 4 ◦C for two days, once the concentration of the solution was equilibrated, the brain sunk. On the left side, various structures were separated, including the hippocampus and the cortex, from which gene expression studies were carried out. 2.9. Brain freezing and cutting Once the brains sunk in sucrose solution, the right hemispheres were frozen in liquid nitrogen and stored at −20 ◦C until they were cut. A cryostat at –22 ◦C was used to obtain the slices, allowing a thickness of 30 μ m per slice. Slices were stored at −20 ◦C in 24-well plates with antifreeze liquid (20% of the total volume of distilled H 2 O, 20% PB2x (Na 2 HPO 4 at 5.73% and NaH 2 PO 4 at 0.624% in distilled H 2 O), 30% ethylene glycol and 30% glycerol). 2.10. Immunofluorescence To assess the involvement of Casp3 in modulating hippocampal neurogenesis, the presence of microglia, their proliferative capacity, and the status of immature neurons in the hippocampus were studied using different primary antibodies as characteristic markers (Table 1) following the protocol previously described by Bachiller et al. (Bachiller et al., 2018). Free-floating hippocampal sections were permeabilized with 1% (v/v) Triton X-100 in PBS for 1 h, and then incubated with 5% (w/v) serum bovine albumin (BSA: Sigma Aldrich) and 1% Triton X-100 in PBS for 1 h to block nonspecific antibody binding. The tissue was then incubated overnight at 4 ◦C with the primary antibodies at the desired dilution in 1% PBS-T and 1% BSA (Table 1). The next day, sections Table 1 Primary antibodies used in immunofluorescence and immunohistochemistry. Antigen Dilution Host Description Commercial house Reference Iba1 1:500 Goat Microglia marker Abcam ab5076 Pro-Caspase-3 1:250 Rabbit Procaspase-3 marker Cell Signaling 9662S Cleaved-Caspase-3 1:100 Rabbit Casp3 activated marker Cell Signaling 9661S Doublecortin (DCX) 1:400 (IF)1:1000 (IHC) Rabbit Immature neuron marker Cell Signaling 4604S Doublecortin (DCX) 1:100 Goat Immature neuron marker Santa Cruz Biotechnology SC-8066 Iba1 1:250 (IF)1:2000 (IHC) Rabbit Microglia marker Wako 01,919,741 Iba1 1:250 Mouse Microglia marker Millipore MABN92 Gal3 1:500 Goat Microglia marker R&D AF1197 GFP 1:500 Goat YFP marker Novus NB100-1770 Table 2 Secondary antibodies used in immunofluorescence and immunohistochemistry. Antigen Dilution Host Commercial House Reference Anti-IgG-Rabbit 488 1:500 Donkey Life Technology A21206 Anti-IgG-Goat 546 1:500 Donkey Life Technology A11056 Anti-IgG mouse 488 1:500 Donkey Life Technology A21202 Anti-IgG Mouse 546 1:500 Donkey Life Technology A10036 Anti-IgG-Rabbit 647 1:250 Donkey Life Technology A31573 Anti-IgG-Rat 488 1:500 Donkey Life Technology A21208 Anti-IgG-Goat 647 1:500 Donkey Life Technology A21447 Anti-IgG-Goat 488 1:500 Donkey Life Technology A11055 Biotinilated anti-IgG mouse mouse 1:250 Goat Vector Laboratories BA-9200 Biotinilated anti-IgG- Rabbit 1:250 Goat Vector Laboratories BA-1000 Biotinylated anti-IgG goat 1:250 Horse Vector Laboratories BA-9500 I.M. Alonso Bellido et al. Brain Behavior and Immunity 112 (2023) 206–219 209 washed for 1 h in PBS containing 0.1% Triton X-100. Then they were incubated for 1 h with the corresponding secondary antibodies (Table 2) in 0.1% PBS-T and 1% BSA for 1 h under agitation and in the absence of direct light and rinsed again with PBS containing 0.1% Triton X-100 for 60 min. For Hoechst staining (1:4000 1 μ g/ml; Sigma Aldrich), the sections were first washed in PBS containing 0.1% Triton X-100 and incubated with Hoechst (1:4000 1 μ g/ml; Sigma Aldrich) for 5 min. The sections were then washed 2x10 min in PBS containing 0.1% Triton X- 100. Brain sections were mounted in 50% glycerol for visualization. All images were acquired under similar conditions (laser intensities and photomultiplier voltages) using a confocal microscope (ZEISS LSM 7 DUO) located at CITIUS, belonging to the University of Seville. Image analysis was performed with free Fiji ImageJ software (W. Rasband, National Institutes of Health). Several parameters were measured with this program, such as the area and perimeter occupied by Iba1 and the fluorescence intensity of Casp3. Furthermore, using DCX-labeled hippocampal slices and Gal3, immature neurons and PAM microglia were counted in the dentate gyrus (DG). For these counts, the ’’Cell counter’’ function of the ImageJ program was used again. Finally, to quantify the percentage of phagocytic microglia, a region of interest (ROI) containing the DG was created for each image. By establishing the same threshold for Iba1 fluorescence on each image, a binary image was created. After a process of erosion, dilation, and watershed, a particle analysis was performed. Thus, a library of ROIs that contained each microglia was created. Phagocytosis was defined as the formation of an enclosed threedimensional pouch of microglial processes that surround an apoptotic cell. Apoptotic cells were defined according to their nuclear morphology after Hoechst staining as cells in which the structure of the chromatin (euchromatin and heterochromatin) was lost and appeared condensed and/or fragmented (pyknosis/karyorrhexis) (Diaz-Aparicio et al., 2020). For surface Imaris reconstruction, Z-stacks at 0.64 μ m (1024x1024) taken using ZEISS LSM 7 DUO confocal laser scanning microscope at 63x, were opened in Imaris (v.9.6, Oxford Instruments) in their original format. Z-stacks were automatically reconstructed into a 3D model, without further image pre-processing. 2.11. Immunohistochemistry First, Iba 1 immunohistochemistry (Table 1) was used for microglial quantification. The entire series of floating brain slices were washed with PBS to remove excess antifreeze. The slices were then incubated for 10 min in 1 ml of MetOH (99% MetOH and 1% H 2 O 2 ) to inactivate endogenous peroxidase in the tissue. The slices then immersed in 2.5% PBS triton X-199 (PBS-T) for tissue permeabilization. To prevent nonspecific binding of the antibody, the sections were incubated for 2 h in 2.5% PBS-T and 5% BSA (Sigma Aldrich). After incubation, the primary antibody was added at the corresponding dilution (Table 1) in 2.5% PBS-T and 2.5% BSA and incubated overnight at 4 ◦C. The next day, sections were kept for 30 min at room temperature and then washed with 2.5% PBS-T. Sections were then incubated with a biotinylated secondary antibody (Table 2) in PBS-T 2.5% and BSA 1% for 2 h at room temperature. After this time, the samples were washed in PBS. The ABC kit (Vector, PK-6100; 1 %A, 1 %B, 98% PBS1x) was then added and kept at room temperature for 1 h. Visualization of immunostaining was achieved using the 3,3 ′ -diaminobenzidine tetrahydrochloride (DAB; Vector, SK-4100) reaction (1 min in 0.2 M Tris buffer containing 0.1% DAB and 0.001% hydrogen peroxide). After the reaction, washes were performed with PBS to remove excess DAB and stop the reaction. The samples were then mounted on gelatin-coated slides and left to dry overnight at room temperature. Once dried, tissue dehydration was performed using a battery of increasing concentration alcohols: 5 min in 70% ethanol, 10 min in 80%, 90%, and 100% ethanol, and 10 min histolemon. Finally, DPX (Scharlau, DP00500500) was added and covered with coverslips for subsequent visualization. Images were acquired with a bright-field microscope (Olympus BX61) located at the CITIUS belonging to the University of Seville. For microglia quantification, the number of microglial cells was counted as a whole using a bright-field microscope with a 40x objective in two regions of the brain, namely the cortex and the hippocampus differentiating CA1 and DG. Each count was performed five times in five regions of interest per section. Three sections per animal were used. 2.12. Stereology To perform a correct analysis of DCX-labeled immunohistochemical slides, stereology was used following the protocol described previously (García-Domínguez et al., 2021). Briefly, the number of DCX-positive neurons in the DG was estimated using a fractional sampling design using CAST software (Olympus Stereology System, Melville, NY, USA) coupled to the Olympus BX61 optical microscope. Immature neurons were counted at regular intervals with predetermined dimensions (x = 150 μ m and y =200 μ m). A grid of known area (141.4x106.0 μ m =1500 μ m 2 ) was superimposed on the counting point and used with the 40x objective. Therefore, the fraction of the sampling area for each of the sections was 1500/(150x200) =0.05 μ m 2 . The Z-axis of each section was normalized; therefore, the thickness of the sampled fraction was 1. Therefore, counting was carried out on 30 μ m slices 360 μ m apart, so the percentage of samples analyzed was 30/360 =0.08. The number of immature DG neurons was calculated by multiplying the number of neurons counted within the counting region by the inverse of the fraction of the sampling area and the inverse of the total number of sections counted. 2.13. RT-PCR The hippocampi and the cortex were extracted from one of the hemispheres of the experimental subjects and preserved at −80 ◦C. The Bioline TRIsureTM protocol (BIO-38033) was followed, which allows the separation of cells and tissues in three phases: DNA, RNA, and proteins. The cDNA was synthesized from 1 μ g of total RNA using the RevertAidTM First Strand cDNA Synthesis kit from Thermo ScientificTM (Thermo Scientific, K1622) in a reaction volume of 20 μ l as described by the manufacturer. Real-time PCR was performed using 5 μ l of SensiFAST ™ SYBR NO-ROX KIT (Bioline, USA BIO-98020), 0.4 μ l of each primer and 4.2 μ l of cDNA to arrive at a final reaction volume of 10 μ l for a 384- well plate. Quantification was performed in triplicate by adding pure water control. Amplification was performed on a Lightcycler® 480 Instrument II thermal cycler (Roche) at 95 ◦C for 2 min followed by 40 cycles consisting of a denaturation step for 5 s at 95 ◦C, a second hybridization step at 65 ◦C for 10 s, and a final elongation step at 72 ◦C for 20 s. The process was completed with a final step of 7 min at 72 ◦C. The analysis confirmed a single PCR product. GAPDH served as a reference gene and was used for sample normalization. The cycle at which each sample crossed a fluorescence threshold (Ct value) was determined and the values were averaged in triplicate for each cDNA. The primer sequences are: Gapdh: FW: GTGTTTCCTCGTCCCGTAGA, RV: AATCTCCACTTTGCCACTG; dcx: FW: CCAATGCGCCGCAGCAAGTC, RV: TGACAGCGGCAGGTACAGGTC; Cx3cR1: FW: GCCTCTGGTGG AGTCTGCGTG, RV: TGGGCTTCCGGCTGTTGGTG; MerTK: FW: CTGTCCAAATCCACAATGCCAC, RV: GTTGACGAGGGTGCGTAATC; Rab7: FW: TGCTGAAGGTCATCATCCTG, RV: AGAGACTG GAACCGTTCTTG; Trem2: FW: GTTTCTTGCAGCCAGCATCC, RV: GGGTCCAGTGAGGATCTGAAG; CD11b; FW: CCTTGTTCTCTTTGATGCAG, RV: GTGATGACAACTAGGATCTT; CD68; FW: GCCCAAGGAACAGAGGAAG, RV: GGCTGTAGGTGTCATCGTG; Axl; FW: GTTGG TGTCTGGAGGATGGG, RV: TGTGTGTCCTTATGGGCTGC; Clec7a; FW: CTGGTATGGAAGTAAGAGACACTGC, RV: CGGTGAGACGATGTTTGGC and P2ry12; FW: GCAGAACCAGGACCATGGAT, RV: CTGAC GCACAGGGTGCTG. I.M. Alonso Bellido et al. Brain Behavior and Immunity 112 (2023) 206–219 210 Fig. 1. Microglia associated with neurogenic niches co-localize with Clv-Casp3. Representative images from the immunofluorescence assay of the DG of the hippocampus (A) of a control mouse. High magnification images confocal (B) and IMARIS image surface representation (C) from the immunofluorescence assay of the DG (SGZ) of the hippocampus using antibodies against DCX, as a new-born neurons marker (in gray), Iba1 as general microglia marker (in red) and Clv-Casp3 (in green). Scale bar 200 µm (A), 20 µm (B) and 5 µm (C). Yellow arrows indicate Iba1/Clv-Casp3 double-labeled microglia in the hippocampal DG. GLC; granular cell layer, SGZ; subgranular zone, H; Hilus. Yellow dashed lines in A indicate different layers of the DG. (D) Measurement of the percentage of Clv-Casp3 labeling in Iba1 + cells in the hippocampus and cortex in control mice. Data are expressed as mean ±SEM of N =4, 4 mice/group. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) I.M. Alonso Bellido et al. Brain Behavior and Immunity 112 (2023) 206–219 211 2.14. Statistical analysis For statistical analysis of the data obtained, the Sigma Plot for Window software (Sigma Plot v11.0 Build 11.0.0.77; SYSTAT, San Jose, CA, USA) was used. Once it was verified that all the data followed a normal distribution and homoscedasticity, two tail Student’s t-test was carried out. Otherwise, non-parametric tests Kruskal-Wallis One Way Analysis of Variance on Ranks were used and indicated in figure legend. The level of significance was set to p <0.05. 3. Results 3.1. Cleaved-Casp3 is present in microglia in cortex and hippocampus brain areas in mice. Our study yielded interesting results, revealing that around 70% of microglia in the cortex and hippocampus expressed Casp3 (Fig. 2C). This led us to question whether activated Casp3 (Clv-Casp3) was present in non-apoptotic microglia in both regions of control mice. To investigate this, we used an immunofluorescence approach and found that approximately 50% of microglia exhibited Clv-Casp3 in both regions (Fig. 1D). Strikingly, we observed that Clv-Casp3 was more prominently present in microglia associated with neurogenic niches (as shown in Fig. 1B and C and Supplementary Video S1). Having established the presence of Casp3 in microglia associated with neurogenic niches, we aimed to further investigate its role in microglia associated with AHN. 3.2. Cx3cr1-CASP3KO conditional mice (Casp3-cKO). To study the role of Casp3 in microglial function, a conditional mouse with Casp3 reduction was generated in these cells in adult mice (see Material and Methods section and Fig. 2A). The clear advantage of this model was that the induction of conditional mice depends on tamoxifen administration; therefore, the development of the mice occurred normally and the Casp3 microglial could be depleted at a specific time point, in our mice model at 2–3 month old (Fig. 2A). First, we wondered if our experimental mice had a significant knockdown-expression of Casp3 in microglia. To this end, we performed immunofluorescence assays to detect Casp3 expression in Iba1 + cells, corresponding to microglia, in conditional mice in hippocampus and cortex (Fig. 2B and Supplementary Fig. S1A). The intensity of the observed Casp3 label showed a significant reduction of ~ 57% (p < 0.05) in the mean intensity of Casp3 in Iba1 + cells in the hippocampus and ~ 68% (p <0.02) in cortex of experimental mice compared to the Fig. 2. Casp3-cKO mice generation and experimental model testing. (A) Caspase-3 f/f C57BL/6 mice with the Casp3 allele floxed in exon 2 were crossed with C57BL/6 mice containing a CRE recombinase under the control of the Cx3cr1 promoter to obtain the experimental Cx3cr1 Cre/WT -CASP3 f/f (Casp3-cKO) and control Cx3cr1 WT/WT -CASP3 f/f mice (Casp3-control). After 3–4 months we induced Casp-3 reduction by tamoxifen injection. A 3 weeks later, we conducted the behavioral test, and finally we proceeded to tissue extraction a 2 weeks later. (B) Immunofluorescence assays to detect Casp3 expression in iba1 + cells, corresponding to microglia, in conditional mice. Scale bar =20 µm. Yellow asterisks indicate Iba1/ Casp3 double-labeled microglia in the hippocampus. (C) Measurement of the percentage of Casp3 labeling in Iba1 + cells in the hippocampus and cortex in control mice. N =5, 5 mice/group. (D) Measurement of the mean intensity of Casp3 labeling in Iba1 + cells in the hippocampus of experimental mice (42,99 ±3,8) compared to the control group (100 ±27,87) showed a significant decrease of ~ 57% (p <0,05; Kruskal-Wallis One Way Analysis of Variance on Ranks). N =4, 4 mice/group. Data are expressed as mean ±SEM. *, p <0.05 versus control mice. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) I.M. Alonso Bellido et al. Brain Behavior and Immunity 112 (2023) 206–219 212 control mice (Fig. 2D and Supplementary Fig. S1B, respectively). This result confirms that the lox-cre system to generate conditional reduction of Casp3 in microglia works in our model. Casp3 reduction in microglia alters the number of the microglia. Once Casp3 reduction in microglia was confirmed after tamoxifen administration, we wanted to know the effect that this reduction had on the microglial population. The first experimental approach was to quantify the number of microglia in brain sections using immunostaining with Iba1 + to detect microglia, specifically in the cortex and hippocampus (including CA1 and GD regions (Fig. 3A). A significant reduction in the number of microglial cells was observed in the conditional animals in the two hippocampal regions studied (CA1: ~16% (p <0.01) and DG: ~26% (p <0.001), Fig. 3D and Fig. 3E respectively) compared to Cre negative littermates. A similarity result was obtained when estimating the density and the total number of microglia (Fig. 3F, G; p <0.004). However, no differences were found in the number of microglia in the cortex (Fig. 3B; P =0.31). We additionally measured mRNA levels of Cx3cr1, a homeostatic gene highly expressed by canonical microglia (Butovsky and Weiner, 2018), being even higher in its expression in the DG (Artegiani et al., 2017). This analysis showed a Fig. 3. Effect of Casp3 reduction on microglia: alteration of the number of the microglia. Immunostaining of Iba (A) for microglia quantification revealed no differences in the cortex in the number of microglia (p =0.31; t-test) (B) and in Cx3cr1 mRNA levels, N =4, 4 mice/group (p =0.351; t-test) (C), but did in the two studied zones of the hippocampus, CA1 (~16%; p <0.01; Kruskal-Wallis One Way Analysis of Variance on Ranks) (D) and GD (~26%; p <0.001; t-test) (E). Scale bar =50 µm. Density (F) and total number of microglia (G) in hippocampus was reduced ~ 18% in conditional mice in comparison of control groups (F and G; p < 0.01; t-test), N =5, 6 mice/group. A ~ 60% reduction in Cx3cr1 mRNA levels was also shown, N =4, 6 mice/group (H; p <0.02; t-test), N =4, 6 mice/group. Data are expressed as mean ±SEM. GCL; Granular cell layer, SGZ; H; Hilus. Blue dashed lines in A indicate different layers of the DG. *, p <0.05; **, p <0.01; ***, p < 0.001 versus control mice. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) I.M. Alonso Bellido et al. Brain Behavior and Immunity 112 (2023) 206–219 213 reduction in Cx3cr1 mRNA expression of ~ 60% in the hippocampus from animals lacking Casp3 (Fig. 3H; p <0.02) but not in cortex (Fig. 3C; p =0.35) thus, together with the quantification of microglia by immunohistochemistry against Iba1 confirming the presence of fewer microglial cells in this region of the brain. Therefore, these data suggest that Casp3 play a role in maintaining hippocampal microglia. As the density of microglial cells in the hippocampus was reduced, we wondered if this was also the case in the neurogenic niches. Recently, several studies have described a microglia subpopulation during development that exhibited a preferential location in white matter and neurogenic niches that host a specific molecular signature. This microglia subpopulation is characterized by the selective expression of different markers including Spp1, Cd11c, Clec7a and Lgals3 and their high phagocytic function (Li et al., 2019; Hammond et al., 2019; Artegiani et al., 2017). The Lgals3 gene encodes galectin-3 (Gal3), a protein expressed by various microglia subtypes both in normal physiological states, such as CD11c + microglia, proliferative-region-associated microglia (PAM), and axon tract-associated microglia (ATMs), as well as in pathological conditions like disease-associated microglia (DAM), microglia neurodegenerative phenotype (MGnD), and white matterassociated microglia (WAM) (Li et al., 2019; Keren-Shaul et al., 2017; Krasemann et al., 2017). Interestingly, homeostatic microglia do not express Gal3 (Butovsky and Weiner, 2018). These characteristics along with a recent described role of Gal3 in neurogenesis (for review, please see (Soares et al., 2021)) made it necessary to study it in our model. First, and importantly, we identified a microglia subpopulation specifically located in the in the subgranular zone (SGZ) of the hippocampus which express Gal3 (Fig. 4A). Next, we wondered if these Gal3 + /Iba1 + positive microglia cells also decrease in number in Casp3-cKO mice compared to controls. Again, representative confocal images clearly showed a decrease in the number of microglial cells (Iba1 + ) in the hippocampus in conditional mice with respect to controls (Fig. 4A, Iba1 staining). Interestingly, the expression of Gal3 in cortex microglia was practically undetectable (Supplementary Fig. S1D). In addition, it can be observed that the Gal3-labeled microglia also decreased ~ 62% in the conditional mouse (Fig. 4A, Gal3 staining and Fig. 4B, C; p <0.001 and p <0.05). Furthermore, ~74% of Gal3 + microglia colocalized with Clv-Casp3 (Fig. 4D, E). Together, these results could indicate that the conditional depletion of Casp3 alters the number of total microglia in the hippocampus and decreases the number of Gal3 + /Iba1 + in the SGZ, where neurogenesis takes place. Casp3-cKO mice show a decrease in new-born neurons in the dentate gyrus of the hippocampus. Microglia is one of the main cellular components that regulates Fig. 4. Effect of Casp3 reduction on Gal3 expression in DG microglia. (A) Representative confocal images labelling NeuN as neuronal marker (in gray), Iba1 as a general microglia marker (in green) and Gal3 as a PAM marker (in red). Yellow arrows indicate Iba1/Gal3 double-labeled microglia in the hippocampal DG. (B) Quantification of percentage Gal3 positive and Iba1 positive microglia in the DG of the hippocampus in control and conditional mice (p <0.001; t-test; N =5, 6 mice/ group). (C) A ~ 62% reduction in the number of Gal3 + microglia was found in conditional mice in comparison with control one (p <0.02; t-test; N =5, 5 mice/ group). (D) High magnification images from the immunofluorescence assay of the (DG) of the hippocampus using antibodies against Gal3, as PAM microglia marker (in red), Iba1 as a general microglia marker (in gray) and Clv-Casp3 (in green). Yellow arrows indicate Iba1/Gal3/Clv-Casp3 triple-labeled microglia in the hippocampal DG. (E) Measurement of the percentage of Clv-Casp3 labeling in Gal3 + cells in the hippocampus and cortex in control mice (N =4, 4 mice/group). Data are expressed as mean ±SEM. GCL; Granular cell layer, SGZ; Subgranular zone, H; Hilus. Yellow dashed lines in A indicate different layers of the DG. ***, p <0.001 versus control mice. Scale bar =20 µm. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) I.M. Alonso Bellido et al. Brain Behavior and Immunity 112 (2023) 206–219 214 hippocampal neurogenesis (see the Introduction section). After reduction of Casp3 in microglia, we found fewer microglial cells close to the neurogenic niche in the hippocampus. We additionally provide evidence that Gal3 specifically marks microglia associated to the hippocampal neurogenic niches, and this labelling was significantly diminished in mice lacking microglial Casp3. This suggests that Casp3 is essential for this process. DG is the hippocampal region where neurogenesis is mainly modulated, so we wondered if the alteration of the microglia in the conditional model in this region alters the neurogenic process. First, using immunofluorescence approach, we quantified the number of DCX + cells per mm 2 (Fig. 5A) and found a decrease in DCX-labeled neurogenic neurons of ~ 32% in conditional mice compared to controls (Fig. 5B; P <0.001). To confirm this observation, the total number of DCX + per septal hippocampus was estimated using a stereological approach (Fig. 5C). The number of new-born neurons decreased ~ 31% in conditional mice (Casp3-control: 53,608 ±2587 DCX + cells, n =6; Casp3-cKO: 36,953 ±3946 DCX + cells, n =7; p <0.001) (Fig. 5D). Furthermore, when we measured the expression of DCX in hippocampal tissue, a decrease in DCX mRNA levels was observed (Fig. 5E; p <0.05). There was no discernible difference in DCX expression in the cortex between the groups studied (Supplementary Fig. S1C). These results were consistent with the decrease in new-born neurons in the conditional mice. All these data indicate that the conditional depletion of Casp3 in the microglia in the mice has a disruptive effect on microglia associated to neurogenic niches in the hippocampus that results in a significant decrease in adult hippocampal neurogenesis. Microglia with reduction of Casp3 display decreased phagocytic capacity. Microglia have been described as the main component of the SGZ responsible for the elimination of excess new-born cells through phagocytosis to maintain homeostasis (Sierra et al., 2010). This statement, together with the fact that our analysis showed less Gal3 expressing microglia in mice lacking Casp3, could indicate a decrease in their phagocytic capacity. To gain insight into this issue, phagocytic Fig. 5. Casp3 reduction caused a decrease in newly born neurons in the dentate gyrus of the hippocampus. Immunofluorescence of DCX in the hippocampus of Casp3-control mice (top image) and in Casp3-cKO (bottom image) (A) Scale bar =100 µm. Quantification of neurogenic neurons labeled with DCX hippocampus showed a decrease of ~ 32% in conditional mice compared to controls (p <0.001; ttest) (B) Data are expressed as the number of DCX + cells per mm 2 . N =3/3 mice/group. Immunohistochemistry of DCX in the hippocampus (C) and stereological analysis (D) of the total number of DCX + cells revealed a decrease of a ~ 31 of neurogenic neurons in conditional mice (Casp3-control: 53,608 ±2587 DCX + cells, n =6; Casp3-cKO: 36,953 ±3946 DCX + cells, n =7; p <0.001; t-test). The measurement of DCX levels by mRNA in the hippocampus also showed a decrease in conditional mice (p <0.05; ttest). N =4, 6 mice/group (E). Data are expressed as mean ±SEM. *, p <0.05; ***, p <0.001 versus control mice. I.M. Alonso Bellido et al.