The impact of astrocyte calcium signaling in cortico-limbic function and behavior
Abstract
Os astrócitos desempenham múltiplas funções desde a homeostasia cerebral ao controlo e processamento da atividade sináptica. Eles integram sinais neuronais por elevações complexas de cálcio (Ca2+) com impacto na comunicação neurónio-astrócito. As elevações de Ca2+ nos astrócitos podem ser divididas em dois tipos: globais (presentes no soma e principais processos) e/ou focais (presentes nos microdomínios). Apesar de estar descrito que as elevações globais de Ca2+ nos astrócitos podem modular a comunicação sináptica, continuam por esclarecer quais os mecanismos moleculares envolvidos. Desta forma, urge uma caracterização comportamental, estrutural e molecular detalhada para compreender esses mecanismos. Nesta tese, utilizámos o modelo de murganho que apresenta a deleção constitutiva do receptor 2 do inositol 1,4,5-trifosfato (IP3R2 KO), no qual as elevações globais de Ca2+ nos astrócitos estão ausentes. Na primeira parte deste trabalho (Capítulo 2) demonstramos que os murganhos IP3R2 KO têm um desenvolvimento somático e neurológico normal. Em seguida, a caracterização comportamental deste modelo transgénico (Capítulo 3) revelou que os murganhos IP3R2 KO apresentam uma melhoria do desempenho cognitivo em tarefas dependentes do hipocampo. Identificámos o factor de transcrição Foxo1 como modulador da expressão de genes específicos de astrócitos, responsáveis pela regulação do citoesqueleto e de espinhas dendríticas. A sobre-expressão do FOXO1 em astrócitos do hipocampo de murganho C57BL/6J foi suficiente para mimetizar a melhoria cognitiva verificada no modelo IP3R2 KO. Este resultado levou-nos a avaliar o papel da sinalização global de Ca2+ no contexto da depressão, uma doença que afeta comportamento dependente das regiões cortico-límbicas (Capítulo 4). Os murganhos IP3R2 KO apresentam uma surpreendente resiliência ao efeito ansiogénico do stress crónico. Por fim, explorámos o papel da sinalização de Ca2+ nos astrócitos no envelhecimento cognitivo (Capítulo 5). Os nossos resultados demonstram uma preservação do desempenho cognitivo em murganhos IP3R2 KO envelhecidos, caracterizado por alteração do rácio neurónio/astrócito e por refinamento dendrítico dos neurónios da camada V do córtex pré-frontal. Em suma, este trabalho contribuiu para uma melhor compreensão do papel da sinalização global de Ca2+ nos astrócitos desde o desenvolvimento até ao envelhecimento, num contexto de saúde e doença. Os resultados revelaram um alvo terapêutico específico em astrócitos com potencial aplicação em contextos de depressão e envelhecimento cognitivo.
Full text
ii DIREITOS DE AUTOR E CONDIÇÕES DE UTILIZAÇÃO DO TRABALHO POR TERCEIROS Este é um trabalho académico que pode ser utilizado por terceiros desde que respeitadas as regras e boas práticas internacionalmente aceites, no que concerne aos direitos de autor e direitos conexos. Assim, o presente trabalho pode ser utilizado nos termos previstos na licença abaixo indicada. Caso o utilizador necessite de permissão para poder fazer um uso do trabalho em condições não previstas no licenciamento indicado, deverá contactar o autor, através do RepositóriUM da Universidade do Minho. Attribution 4.0 International (CC BY 4.0) https://creativecommons.org/licenses/by/4.0/
iii Agradecimentos/Acknowledgments Chegada ao final desta caminhada, quero agradecer a todos aqueles que, no momento certo, me souberam dar uma palavra de aconchego e de orientação, força e confiança para que este trabalho chegasse a bom porto. Ao Doutor João Filipe Oliveira por me ter contagiado com a sua paixão pelo mundo dos astrócitos! Obrigada pela partilha de conhecimentos, pela orientação, pela boa disposição e pelo entusiasmo! À Doutora Luísa Pinto por ter aceite ser minha orientadora! Obrigada pelas sugestões, pelas discussões científicas e pelo apoio constante! Aos Professores Nuno Sousa e João Bessa, pelo trabalho desenvolvido como coordenadores dos NERDs. Aos Professores Fernando Rodrigues, Margarida CorreiaNeves e Joana Palha pela confiança que depositaram em mim como Representante dos alunos dos anos avançados do PhDCS. Aos Professores Jorge Pedrosa e Jorge Correia Pinto por trabalharem diariamente no sentido de nos proporcionarem as melhores condições de trabalho, mesmo em dias menos risonhos. A todos os funcionários do ICVS e da Escola de Medicina, com especial agradecimento às tratadoras do Biotério e à Doutora Alice Miranda. À Doutora Teresa Summavielle por ter aceite ser membro externo da Comissão que acompanhou esta tese. À Doutora Andreia Teixeira-Castro pelo apoio no trabalho experimental, pelas discussões, por ser um exemplo de cientista a seguir! Ao meu Astrogang, atuais membros: Inês “a engenheira”, João Viana “o vírus” e Diana “a poveirinha”; e anteriores membros: Gabriela, Manuella, Daniel, Pedro Pinto, Joana Correia, Ana Oliveira e Ana Lima. À Vanessa, que foi uma das maiores conquistas deste doutoramento! Ganhei uma irmã para a vida, e se cheguei até aqui muito te devo a ti! Aos melhores vizinhos de secretária ever: Bárbara Pinheiro, Sofia Serra e Eduardo Gomes. Aos amigos que este doutoramento me deu: Cláudia Antunes, Luísa Santa Marinha, Jorge Silva, Marco Guimarães, Sofia Neves, Susana Monteiro, Dulce Almeida, Rita Silva, Marta Guedes, Isabel Castanho, Sara Pinheiro, Fátima Ramalhosa, Filipe Marques, João Costa, André Lopes, Mónica Morais e Francisca Bravo. Ao Eduardo Campos e ao Dinis Alves por todas as vezes que me ajudaram no biotério. Aos meus amigos de Aveiro: Ana Macedo, Carla Pereira, João Carvalho, Sara Marques e Tamira Cruz. Aos alcoólicos anónimos: Sílvia e Diogo, Samuel e Sofia, Rui e Tânia, Andreia e Daniel. Aos meus amigos de Braga: Carla, César e Francisca, Diogo, Manuel, José Antunes e Dominique. Ao meu afilhado Francisco. À minha família da Póvoa e Cabeceiras, em especial avós, tios e primos. Aos meus pais e ao meu irmão, que sempre acreditaram em mim e me deram todo o amor, força e confiança que precisei ao longo da vida. Ao Pedro, sem ti nada disto seria possível! Juntos somos mais fortes e a chegada ao final desta etapa é a prova disso. Esta tese é tua! The work presented in this thesis was performed in the Life and Health Sciences Research Institute (ICVS), at the School of Medicine, University of Minho. Financial support was provided by a PhD grant (SFRH/BD/101298/2014 to SGG), FCT Investigator grants (IF/00328/2015 to JO, IF/01079/2014 to LP) and PTDC/MED-NEU/31417/2017 from the FCT – Foundation for Science and Technology, by BIAL Foundation grants (207/14 to JO and 427/14 to LP), by Northern Portugal Regional Operational Programme (NORTE 2020), under the Portugal 2020 Partnership Agreement, through the European Regional Development Fund (FEDER) (NORTE-01-0145-FEDER-000013); FEDER Funds, through the Competitiveness Factors Operational Programme (COMPETE), and The National Fund, through the FCT (POCI-01-0145-FEDER-007038).
v STATEMENT OF INTEGRITY I hereby declare having conducted this academic work with integrity. I confirm that I have not used plagiarism or any form of undue use of information or falsification of results along the process leading to its elaboration. I further declare that I have fully acknowledged the Code of Ethical Conduct of the University of Minho.
v O impacto da sinalização de cálcio nos astrócitos na função cortico-límbica e no comportamento Resumo Os astrócitos desempenham múltiplas funções desde a homeostasia cerebral ao controlo e processamento da atividade sináptica. Eles integram sinais neuronais por elevações complexas de cálcio (Ca2+) com impacto na comunicação neurónio-astrócito. As elevações de Ca2+ nos astrócitos podem ser divididas em dois tipos: globais (presentes no soma e principais processos) e/ou focais (presentes nos microdomínios). Apesar de estar descrito que as elevações globais de Ca2+ nos astrócitos podem modular a comunicação sináptica, continuam por esclarecer quais os mecanismos moleculares envolvidos. Desta forma, urge uma caracterização comportamental, estrutural e molecular detalhada para compreender esses mecanismos. Nesta tese, utilizámos o modelo de murganho que apresenta a deleção constitutiva do receptor 2 do inositol 1,4,5-trifosfato (IP3R2 KO), no qual as elevações globais de Ca2+ nos astrócitos estão ausentes. Na primeira parte deste trabalho (Capítulo 2) demonstramos que os murganhos IP3R2 KO têm um desenvolvimento somático e neurológico normal. Em seguida, a caracterização comportamental deste modelo transgénico (Capítulo 3) revelou que os murganhos IP3R2 KO apresentam uma melhoria do desempenho cognitivo em tarefas dependentes do hipocampo. Identificámos o factor de transcrição Foxo1 como modulador da expressão de genes específicos de astrócitos, responsáveis pela regulação do citoesqueleto e de espinhas dendríticas. A sobre-expressão do FOXO1 em astrócitos do hipocampo de murganho C57BL/6J foi suficiente para mimetizar a melhoria cognitiva verificada no modelo IP3R2 KO. Este resultado levou-nos a avaliar o papel da sinalização global de Ca2+ no contexto da depressão, uma doença que afeta comportamento dependente das regiões cortico-límbicas (Capítulo 4). Os murganhos IP3R2 KO apresentam uma surpreendente resiliência ao efeito ansiogénico do stress crónico. Por fim, explorámos o papel da sinalização de Ca2+ nos astrócitos no envelhecimento cognitivo (Capítulo 5). Os nossos resultados demonstram uma preservação do desempenho cognitivo em murganhos IP3R2 KO envelhecidos, caracterizado por alteração do rácio neurónio/astrócito e por refinamento dendrítico dos neurónios da camada V do córtex pré-frontal. Em suma, este trabalho contribuiu para uma melhor compreensão do papel da sinalização global de Ca2+ nos astrócitos desde o desenvolvimento até ao envelhecimento, num contexto de saúde e doença. Os resultados revelaram um alvo terapêutico específico em astrócitos com potencial aplicação em contextos de depressão e envelhecimento cognitivo. Palavras-chave: astrócito, cálcio, cérebro, comportamento, IP3R2
vi The impact of astrocyte calcium signaling in cortico-limbic function and behavior Abstract Astrocytes are responsible for distinct functions ranging from brain homeostasis to the modulation of synaptic functioning. They integrate neuronal signals by complex calcium (Ca2+) elevations that control intracellular mechanisms that in turn drive the neuron-astrocyte dialogue, modulating the activity of cells and networks. It is now recognized that Ca2+ elevations in astrocytes appear spatially distributed in global (soma and main processes) and/or focal regions (microdomains). Although it is observed that global astrocytic Ca2+ signaling contributes to synaptic communication, its role in circuit computation and behavioral performance is still poorly understood. A detailed behavioral, structural and molecular characterization should provide us with putative mechanisms underlying the roles of astrocytic Ca2+. In this thesis, we took advantage of the inositol 1,4,5-trisphosphate receptor type 2 knockout (IP3R2 KO) mouse model, which lacks global Ca2+ signaling in astrocytes. In the first part of this work (Chapter 2), we demonstrate that IP3R2 KO mice retain a normal developmental maturation, as compared with WT littermates. Next, a detailed behavioral characterization of this mouse model (Chapter 3) showed that IP3R2 KO mice display enhanced cognitive performance in hippocampal-dependent tasks. We found Foxo1 as the most active transcription factor controlling the increased expression of astrocyte-specific genes related with fine cytoskeleton modulation and spinogenesis, which could underlie the cognitive enhancement observed. Moreover, specific overexpression of FOXO1 in hippocampal astrocytes of C57BL/6J mice was enough to recapitulate the enhanced fear memory observed in IP3R2 KO mice. This striking observation prompted us to test the role of global Ca2+ signaling in the context of depression, which affects cortico-limbic regions (Chapter 4). IP3R2 KO mice present an unexpected resilience to the installation of stress effects, namely translated into an increased self-care and a reduced anxious-like phenotype. Finally, we explored the role of astrocytic Ca2+ signaling in cortico-limbic performance in aged mice that display cognitive decline (Chapter 5). We observed a preserved cognitive performance in aged IP3R2 KO mice, an altered neuron/astrocyte ratio and a dendritic refinement of mPFC neurons. Overall, this work contributed to a better understanding on the role of global astrocytic Ca2+ signaling from development to aging, both in a health and disease context. We found a putative astrocyte-specific therapeutic target that could be used to prevent depressionand aging-related deficits. Keywords: astrocyte, behavior, brain, calcium, IP3R2
vii Contents Agradecimentos/Acknowledgments .................................................................................................... iii Resumo............................................................................................................................................... v Abstract.............................................................................................................................................. vi Abbreviations list ................................................................................................................................ ix Figures and tables list ........................................................................................................................ xiii CHAPTER 1 ...................................................................................................................................... 15 Introduction ...................................................................................................................................... 15 1. Glial cells in the brain ............................................................................................................... 16 Astrocytes ......................................................................................................................... 17 Developmental origin and heterogeneity ..................................................................... 18 Homeostatic functions ............................................................................................... 20 The tripartite synapse ................................................................................................ 22 Astrocytic calcium signaling ............................................................................................... 24 Mechanisms that trigger astrocytic calcium elevations ................................................ 25 Different forms of Ca2+ elevations in astrocytes: the complexity of simplicity .................. 28 Functional roles of astrocyte calcium elevations .......................................................... 29 The IP3R2 KO mouse model ....................................................................................... 36 Cognition: learning and memory ........................................................................................ 42 The hippocampus ...................................................................................................... 44 The prefrontal cortex .................................................................................................. 45 Behavior tests to assess cognition in mouse models ................................................... 45 Astrocytic modulation of cognitive behavior ................................................................. 47 Astrocytes in disease: involvement in depression ............................................................... 49 Astrocytes in aging ............................................................................................................ 50 Aims of this thesis ............................................................................................................. 52
xiv Figure 4.2 – Exposure to uCMS induces reduction of body weight and deteriorates self-care in WT, but not IP3R2 KO mice. ......................................................................................................................... 114 Figure 4.3 – Assessment of anxiousand depressive-like behavior in WT and IP3R2 KO mice exposed to uCMS and respective non-stressed controls. ................................................................................... 115 Figure 5.1 - Lack of IP3R2-dependent astrocytic calcium prevents age-related cognitive decline in a PFCdependent task. .............................................................................................................................. 128 Figure 5.2 - Aged IP3R2 KO mice display decreased NeuN+ neuron densities and increased S100β+ astrocyte densities. ......................................................................................................................... 129 Figure 5.3 - Aging leads to a dendritic refinement of mPFC layer V pyramidal neurons in IP3R2 KO mice. ...................................................................................................................................................... 131 Figure 6.1 – Scheme depicting the behavior characterization performed in IP3R2 KO mice throughout the lifespan. ......................................................................................................................................... 135 Figure S 3.1 - IP3R2 KO mice display neither anxiousnor depressive-like behavior. .......................... 104 Figure S 3.2 – GFAP-mCherry-FOXO1 mice do not display an anxious-like phenotype and present normal locomotor and exploratory abilities. ................................................................................................. 105 Figure S 4.1 - Exposing WT and IP3R2 KO mice to an uCMS protocol do not have an impact in cognitive performance assessed by the Novel Object Recognition test. ........................................................... 120 Table 1.1 - Hallmarks of distinct Ca2+ responses present in the complex astrocytic morphology........... 27 Table 1.2 - Functional consequences of the interference with IP3R2 expression .................................. 38 Table 3.1 - Forward and reverse sequences of oligonucleotide primers of the selected genes used for microarrays data validation. .............................................................................................................. 81 Table 3.2 – iRegulon results regarding the most enriched transcription factors from our list of differentially expressed genes. .............................................................................................................................. 91 Table 3.3 – Fold change, cellular specificity and reported functions of genes in the identified cluster. . 93 Table Annex 1 - List of differentially expressed genes in total hippocampus of IP3R2 KO mice ........... 172
15 CHAPTER 1 Introduction
16 1. Glial cells in the brain The brain is the most complex organ, composed by a diversity of cell types organized in a defined structure. These unique characteristics allow it to regulate and dynamically respond to environmental changes and to coordinate several biological responses. During many years, neurons were recognized as the main cellular elements in the brain, responsible for transmitting and processing information through electrical impulses. Neurons are well organized into circuits and communicate by specialized structures called synapses. One hundred and sixty years ago, Rudolf Virchow revolutionized the neuroscience field by introducing the term “neuroglia”, as non-neuronal cells of the central nervous system (CNS). During many years, glial cells were considered to play a passive role in brain functioning, mainly due to the absence of electrical excitability. They were described as a connective material responsible for filling the interstitial space and for playing a supportive role (Kettenmann and Verkhratsky 2008; McIver et al. 2013). However, several studies revealed that glial cells are not only important for homeostatic brain functions, but they are also active partners in the modulation of synaptic function (Perea et al. 2014). This fact was demonstrated due to the development of calcium (Ca2+) dyes, which uncovered the observation that glial cells respond to the surrounding activity by elevating its intracellular Ca2+ levels (Cornell-Bell et al. 1990; Charles et al. 1991). Several reports described differences in the number, form, and function of glial cells pointing to an increase in the glia to neuron ratio in phylogenetically evolved species (Verkhratsky and Parpura 2014; Verkhratsky and Nedergaard 2016). Glial cells are a heterogeneous population that can be distinguished according to their morphology and associated functions. Microglia act as the immune and phagocytic cells of the brain, oligodendrocytes and Schwann cells play a role in axonal myelination, NG2-glia are acknowledged for their role as progenitor cells and astrocytes are the most abundant glial cell type population in the brain and the main focus of this thesis (Verkhratsky 2006; Jakel and Dimou 2017). A growing body of evidence has shown that astrocytes can sense (be activated by several signals), integrate (elevate its intracellular Ca2+ levels) and respond (release of gliotransmitters) to neuronal activity (Araque et al. 1999; Araque et al. 2014; Perea et al. 2014; GuerraGomes et al. 2017). The main aim of this thesis was to assess the impact of astrocytic Ca2+ signaling in cortico-limbic function and behavior. Throughout this chapter, we will review the main topics related to astrocytic functions, mainly focusing on intracellular Ca2+ signaling and its role in the health and disease conditions.
17 Astrocytes Astrocytes are the most numerous glial cell type in the CNS. These cells were termed astrocytes by Michael von Lenhossek in the 1890s due to their star-shaped morphology (Verkhratsky et al. 2016a). They correspond to a heterogeneous population that can be classified into protoplasmic astrocytes of the grey matter and fibrous astrocytes of the white matter. Moreover, in distinct brain regions, astrocytes display specific structural and functional properties as an adaptation to local circuit requirements, such as the Müller cells in the retina and Bergmann glia in the cerebellum (Emsley and Macklis 2006; Matyash and Kettenmann 2010; Oberheim et al. 2012; Verkhratsky and Parpura 2014; Ben Haim and Rowitch 2017; Farmer and Murai 2017; Mederos et al. 2018). Astrocytes are responsible for distinct functions from brain homeostasis to the control and processing of synaptic functioning. These glial cells regulate brain homeostasis by interacting in close proximity to neuronal cell bodies, dendrites, axons and synapses (Perea et al. 2009). They interact with blood vessels by enwrapping endothelial cells and pericytes with their endfeet, thus controlling neurovascular coupling (McIver et al. 2013). Astrocytes occupy a unique territory without a structural overlap between neighbor cells, in several brain regions. This morphological organization favors a locally and specific astrocytic response, which is important to respond and regulate the activity of a given brain region (Ben Haim and Rowitch 2017). Due to their spatial organization in individual territories, astrocytes display an efficient cell-to-cell communication mechanism through gap junctions, forming a syncytium. These gap junction channels are formed by connexins (Cxs), with Cx30 and Cx43 being the most expressed in astrocytes (Nagy et al. 1999; Giaume et al. 2010; Pannasch and Rouach 2013). These channels allow the diffusion of several small molecules (up to 1.5 kDa) between neighboring astrocytes, such as ions, neurotransmitters and energy metabolites. (Pannasch et al. 2011; Pannasch and Rouach 2013). Astrocytes may also communicate with vicinal astrocytes by generating intercellular Ca2+ waves through ATP release that will act on Gq-G-protein-coupled receptors (GPCR) present in neighboring cells, eliciting a Ca2+ response (Bowser and Khakh 2007). Thus, instead of responding to the surrounding stimuli by generating action potentials, astrocytes display a Ca2+- and Na+- based excitability. Besides Ca2+ elevations, the intracellular concentration of Na+ in astrocytes can be regulated by plasmalemmal channels, ATP-dependent pumps and ion exchangers (Parpura and Verkhratsky 2012). Astrocytes express several membrane receptors, cell surface molecules, voltageand ligand-gated ion channels. However, these ion channels are poorly expressed and consequently they do not present membrane excitability (Araque and Navarrete 2010). Astrocytes broadly express inwardly-retifying potassium (K+) channel 4.1 (Kir4.1) and their membrane is highly permeable to this ion. These cells have
18 the ability to buffer excess extracellular K+ resultant from neuronal activity and to distribute it to sites where its concentration is lower (Seifert et al. 2018). Briefly, astrocytes are actively involved in brain homeostasis and participate in information processing by forming a network that allows an orchestrated response specific for each brain region and adapted to a physiological or pathological state (Araque 2008; Wang and Bordey 2008; Ben Haim and Rowitch 2017). Moreover, it was recently demonstrated that the developmental origin of astrocytes determines its organization as an intercellular network (Gutierrez et al. 2019). Developmental origin and heterogeneity The complexity of the adult CNS generation forecasts a diversity of mechanisms that give rise to different cell types during development. Several extrinsic developmental signals and intrinsic signaling cascades determine cell fate and consequently the generation of distinct neuronal cells with specific functional, morphological and molecular characteristics (Edlund and Jessell 1999; Molofsky et al. 2012). Astrocytes have been described as being as heterogeneous as neurons, with unique properties in different brain regions. Interestingly, a growing body of evidence showed that neurons and glial cells are generated in a temporarily distinct, yet overlapping pattern. In fact, the CNS is originated from the neuroepithelium, which consists of a thin monolayer of epithelial cells – the neural plate. From these neuroepithelial (NEP) cells, radial glia (RG) cells emerge (Kettenmann and Ransom 2005). The progenitor domains associated with both NEP and RG cells are determined by several environmental signals, indicating positional information about both dorsal-ventral and anterior-posterior axes. Among these signals are members of the dorsal bone morphogenetic protein and Sonic hedgehog (Kessaris et al. 2008; Bayraktar et al. 2014). During mammalian CNS development, neural stem cells are primarily generated, followed by glial cells (Figure 1.1). This allows the adjustment of glial cell number and organization into the newly formed neuronal circuit (Miller and Gauthier 2007; Kessaris et al. 2008). By the onset of neurogenesis, RG cells begin to express several astrocytic markers such as vimentin, glial fibrillary acidic protein (GFAP), glutamate aspartate transporter (GLAST) and glutamate transporter 1 (GLT-1) and glutamine synthetase (GS) (Dimou and Gotz 2014). In the mammalian CNS, gliogenesis begins in later periods of embryonic development and continues until late postnatal period. More specifically, astrocytes are generated around embryonic day 18, and their proliferation is almost complete at early postnatal stages (Miller and Gauthier 2007; Wang and Bordey 2008). However, their maturation continues throughout development and coincides with the establishment of synaptic and vascular connections, where astrocytes exert an
19 important role. During this phase, the expression of astrocytic markers such as GFAP, Aquaporin-4, and S100β significantly increases (Molofsky et al. 2012). Several reports indicate that astrocytic heterogeneity is due to their derivation from regionally patterned RG cells (Bayraktar et al. 2014). Astrocytes display a regional morphological and genetic profile that was shown to be linked to their location and function within the CNS (Ben Haim and Rowitch 2017). They also present intrinsic functional features according to specific brain regions that allow them to build a concerted response to network changes. For instance, astrocytes from distinct brain regions exhibit distinct electrophysiological responses (Chai et al. 2017). Likewise, during running activity, Bergmann glia and neocortical astrocytes display contrasting Ca2+ responses (Nimmerjahn et al. 2009). Furthermore, in the rodent cortex, a stratification of Ca2+ responses is observed among the different cortical layers (Takata and Hirase 2008). More so, Cai et al. (2007) demonstrated that in the absence of a transcription Figure 1.1 - The developmental origin of astrocytes. Schematic drawing depicting several phases associated with preand postnatal astrocytic development. Radial glial cells develop from early neuroepithelial cells, which initially generate neurons, and at later stages give rise to glial-committed precursor cells to produce astrocytes. These precursors proliferate and can generate immature oligodendrocytes and immature fibrous or protoplasmic astrocytes. At the postnatal stage, these cells reach a mature state into oligodendrocytes, fibrous and protoplasmic astrocytes, which coincides with important phases of development like synaptogenesis and angiogenesis. During the late postnatal stage, astrocytes also participate in other important functions such as neuronal migration and synapse elimination. Adapted from Molofsky et al. (2012).
20 factor – Olig2 – the formation of white matter astrocytes is compromised, but gray matter astrocytes remain intact, pointing to a differential mechanism related with the generation of both types of astrocytes. Recently, it was shown that the developmental cell lineage also determines the establishment of a gap junctional network between sibling astrocytes (Gutierrez et al. 2019). A growing number of markers have been helping the study of both morphological and molecular profiles of astrocytes. The first molecular marker described for astrocytes was GFAP, a hallmark cytoskeletal intermediate filament. GFAP is composed by 10 isoforms (α, β, γ, GFAPΔexon6, GFAPΔ164, GFAPΔexon7, GFAPΔ135, δ, κ, and ζ) and astrocytes differentially express these isoforms (Kamphuis et al. 2012; Hol and Pekny 2015). Several reports suggest caution for the use of this astrocytic marker, mainly due to astrocyte molecular diversity. In fact, some astrocytes do not stain for GFAP but are positive for other astrocytic markers such as S100β, GS or aldehyde dehydrogenase 1 family member L1 (Aldh1L1). The main reason behind this fact is the GFAP differential expression across brain regions. For example, most hippocampal astrocytes are GFAP-positive, while few thalamic astrocytes are detected by GFAP immunostaining (Sofroniew and Vinters 2010; Khakh and Sofroniew 2015). Similarly, an almost undetectable immunoreactivity of GFAP in the dorsal and ventral mouse cortex was described (Zhang et al. 2019). Other cell types in the nervous system (central and peripheral) may display positive staining for GFAP. In the CNS, detection of GFAP staining is also observed in RG cells in late embryogenic stages, while in the peripheral nervous system non-myelinating Schwann cells were found to be GFAP-positive (Khakh and Sofroniew 2015; Yang and Wang 2015). Regarding morphology, it is important to note that GFAP only labels astrocyte main processes, which covers around 15% of the astrocyte total area (Khakh and Sofroniew 2015). However, the reconstruction of GFAP-stained structures allows a fast and reliable assessment of gross structural alterations in astrocytes (Tavares et al. 2017). Technical advances in the field have allowed unveiling a diversity of molecular markers for astrocytes, which have proven to be pivotal for the development of new genetic targeting strategies (Cahoy et al. 2008; Zhang et al. 2014; Boisvert et al. 2018). Homeostatic functions Several lines of evidence point to astrocytes as the multitasking cells of the brain. Astrocytes are involved in the regulation of several CNS functions, from development to adulthood (Allen and Lyons 2018). They actively participate in brain homeostatic processes such as maintenance of the blood-brain barrier (BBB), control of local cerebral blood flow, angiogenesis, modulation of the extracellular matrix, neurotransmitter
21 uptake, extracellular ion buffering and metabolic support (Wang and Bordey 2008; Araque and Navarrete 2010). Astrocytes directly contact with CNS microvasculature by extending long processes and enwrapping them with their endfeet. This close interaction allows the control of BBB integrity and modulates neurovascular coupling through the release of several gliotransmitters (Alvarez et al. 2013). This strategic location allows astrocytes to supply neuronal cells with nutrients (Allen and Lyons 2018). Astrocytes contribute to vasculature formation by promoting angiogenesis. A good example of this contribution is observed in the retina. More specifically, the platelet-derived growth factor released by retinal neurons triggers the astrocytic release of vascular endothelial cell growth factor that in turn stimulate blood vessel growth (Wang and Bordey 2008). Furthermore, astrocytes can shape the extracellular matrix by promoting neurite growth and contributing to extracellular matrix degradation and remodeling. Specifically, they express a wide range of extracellular matrix and adhesion molecules, such as N-cadherin and matrix metalloproteinases (Wang and Bordey 2008). Astrocytes secrete several growth factors, such as the brainderived neurotrophic factor, glial cell line-derived neurotrophic factor, nerve growth factor, and basic fibroblast growth factor. These growth factors are fundamental to maintain cell viability, oxidantantioxidant balance, metabolic functioning and gene expression control (Cabezas et al. 2016). Besides their role in cell support, astrocytes can control neuronal excitability. They clear the byproducts of metabolism (e.g. ammonium, NH4+) and neurotransmitters (e.g. glutamate, gamma-aminobutyric acid (GABA), dopamine, noradrenaline) released to the synaptic cleft through the expression of protein transporters, predominantly GLT-1/EAAT2 and GLAST/EAAT1. Under physiological conditions, glutamate uptake occurs through an electrochemical gradient (3 Na+/1H+ in exchange for 1 K+), which alters astrocyte membrane potential (Kreft et al. 2009; Allen 2014; Weber and Barros 2015). The expression levels of glutamate transporters are highly dynamic, dependent on the level of neuronal activity (Benediktsson et al. 2012; Allen and Lyons 2018). Conversely, GLT-1 ablation in mice leads to increased levels of extracellular glutamate, and consequently to spontaneous seizure episodes (Tanaka et al. 1997; Allen 2014). Another important astrocytic housekeeping function is the extracellular K+ buffering. The major K+ buffer present in astrocytes is Kir4.1. During periods of intense neuronal activity, K+ accumulation leads to enhanced neuronal excitability and consequently to seizures. In line with these observations, mice lacking Kir4.1 specifically in astrocytes have an accumulation of extracellular K+, impaired glutamate uptake and develop seizures due to increased excitotoxicity (Djukic et al. 2007; Allen 2014). Additionally, another class of K+ channels (TREK-1 and TREK-2) was reported to be present in astrocytes and to contribute to K+ buffering (Rivera-Pagan et al. 2015).
22 Interestingly, astrocytes not only regulate extracellular neurotransmitter concentration but also control its recycling. For example, glutamate is taken up by transporters and metabolized via the glutamateglutamine shuttle. This mechanism leads to the production of glutamine via GS action. Glutamine is redistributed to glutamatergic/GABAergic neurons so that the neuronal production of glutamate/GABA is ensured (Wang and Bordey 2008; Allen 2014). In the same way, astrocytes are important providers of precursor molecules for neurons. For example, they are enriched in lipid synthesis pathways and produce molecules such as, for instance, cholesterol that is essential for presynaptic function maintenance (Allen 2014). Moreover, astrocytes uptake glucose from the blood stream and produce lactate via two pathways: glycogenolysis and glycolysis. Lactate is formed by astrocytes as a response to neuronal activity. The transport of lactate from astrocytes to neurons occurs through monocarboxylate transporters, fulfilling the neuronal energetic needs and providing signals that modulate synaptic function, the astrocyte-neuron lactate shuttle. For instance, in the hippocampus, astrocytic lactate release is essential to memory formation (Suzuki et al. 2011; Magistretti and Allaman 2018). The tripartite synapse The tripartite synapse concept emerged to describe the bidirectional communication between neurons and astrocytes in brain function control (Araque et al. 1999). This view completely challenged the classically accepted paradigm claiming that brain function relies exclusively on neuronal activity. In fact, astrocytes are intimately close to synapses and it is described that one astrocyte is able to make contact with over 100.000 synapses (Bushong et al. 2002). The tripartite synapse concept is based on three main astrocytic features: (i) expression of functional neurotransmitter receptors (e.g. for glutamate, ATP, GABA); (ii) ability to process complex Ca2+ signaling and communicate with vicinal astrocytes, through “Ca2+ waves” and (iii) release of active substances called “gliotransmitters” (e.g. glutamate, D-serine, ATP, GABA) to feedback to neurons (Figure 1.2).
23 Figure 1.2 - The tripartite synapse. Astrocytes closely enwrap preand postsynaptic terminals. They express several neurotransmitter receptors that allow them to sense and respond to neuronal activity. Neurotransmitters released to the synaptic cleft activate astrocytes, through binding to the receptors present on their membrane. Consequently, astrocytic intracellular Ca2+ levels increase. This Ca2+ rise in astrocytes may trigger the release of several substances, named “gliotransmitters”, such as glutamate, ATP, and D-serine, which feedback to neurons to modulate network activity. Furthermore, astrocytes actively participate in other brain functions such as synaptogenesis, regulation of presynaptic function and modulation of the postsynaptic neuron response to neurotransmitters. Adapted from Allen and Barres (2009). Astrocytes sense neuronal activity through a wide variety of neurotransmitter receptors, ion channels and transporters expressed in their membranes (Pannasch and Rouach 2013; Verkhratsky and Parpura 2014; De Pitta et al. 2016). These receptors can be activated by neurotransmitters, “gliotransmitters” or volume transmitters, which consequently trigger astrocyte excitability (Verkhratsky 2009). Astrocytes express a large number of GPCRs for different neurotransmitters such as glutamate, GABA, purines, norepinephrine, dopamine, and cannabinoids (Verkhratsky and Parpura 2014). Among those, activation of a GPCR coupled to a Gq-protein stimulates phospholipase C (PLC), leading to the production of inositol 1,4,5-trisphosphate (IP3), which, in turn, will activate IP3 receptors (IP3Rs) present in the endoplasmic reticulum (ER). Consequently, Ca2+ is released from astrocytic internal stores and its intracellular Ca2+
30 Accordingly, a recent study has showed that acute blockade of Ca2+-dependent IP3 mechanisms impairs LTP, but that this can be rescued by exogenous D-serine (Sherwood et al. 2017). Astrocytic Ca2+ elevations appear to also promote alternative forms of plasticity in the brain. In the hippocampus, cholinergic afferents from the medial septum modulate CA1 synaptic plasticity via an astrocytic Ca2+-dependent mechanism that triggers glutamate release and consequent activation of mGluRs in neurons (Navarrete et al. 2012). This mechanism of synaptic modulation is not restricted to the hippocampus since astrocytic Ca2+ elevations also mediate muscarinic acetylcholine receptordependent plasticity in the somatosensory cortex (Takata et al. 2011), as it will be discussed in the next chapter. Moreover, a recent study showed that the cholinergic input to astrocytes, driven by active/sleep phases, controls Ca2+-dependent D-serine release, ultimately gearing up N-methyl-D-aspartate receptor Figure 1.4 - Excitatory and/or inhibitory signals trigger Ca2+ elevations in astrocytes and lead to gliotransmitter release. Scheme depicting input signals that trigger astrocyte Ca2+ elevations and respective transmitter release. Both excitatory (+) and inhibitory (–) signals cause global or focal Ca2+ elevations in astrocytes (Left), and precede gliotransmitter release that might exert excitation or inhibition of neighboring synapses (Right). For each reference, the region studied is indicated in black (Amy, Amygdala; BrSt, brainstem; CA1, CA1 subfield of the hippocampus; Ctx, cortex; DG, dentate gyrus; hHip, human hippocampus; Str, Striatum). ∗Indicates the studies that described functional consequences to focal Ca2+, rather than global Ca2+ responses.
31 (NMDA) receptor activation at CA1 synapses (Papouin et al. 2017). In a different hippocampal subfield, the cholinergic activation of hilar astrocytes triggers intracellular Ca2+ elevations that precede the activation of hilar inhibitory interneurons, causing a long-lasting GABAergic inhibition of dentate granule cells (Pabst et al. 2016). Further evidences point out that astrocytic Ca2+ mediates endocannabinoid-dependent plasticity. Navarrete and Araque (2008) reported that the activation of astrocytic cannabinoid type 1 receptors (CB1R) leads to intracellular Ca2+ elevations, triggering glutamate release and slow inward currents (SICs) in vicinal pyramidal neurons. Furthermore, endocannabinoid-triggered elevation of astrocytic Ca2+ levels contributes to heteroneuronal LTP (Gomez-Gonzalo et al. 2015). In the striatum, a circuit-specific astrocyte-neuron signaling takes place in which the release of endocannabinoids promotes Ca2+ elevations in a specific population of astrocytes, triggering the release of glutamate that modulates excitability and synaptic transmission (Martin et al. 2015). Finally, in the medial central amygdala, astrocytes receive endocannabinoid signaling to release ATP and depress excitatory synapses from basolateral amygdala via A1 adenosine receptor activation, and enhance inhibitory synapses from the lateral subdivision of the central amygdala via A2A receptor activation. This redundancy system results in medial central amygdala neuronal inhibition with impact in animal behavior (Martin-Fernandez et al. 2017). Interestingly, astrocytic Ca2+ elevations are not exclusive to the action of classic excitatory transmitters. Activation of GABA receptors in astrocytes was shown by different research groups to trigger Ca2+ elevations with functional synaptic consequences. Specifically, GABAergic heterosynaptic depression in the hippocampal CA1 subfield requires astrocytic Ca2+ elevations and ATP release, whose metabolite adenosine will activate A1 receptors (Serrano et al. 2006). More recently, Mariotti et al. (2016) showed that the activation of GABAB receptors evoked somatic Ca2+ elevations, consequently leading to the occurrence of SICs in cortical pyramidal neurons. Curiously, these effects were absent in IP3R2 KO mice suggesting a relationship between Gi-coupled GPCRs and IP3 signaling in astrocytes. Moreover, the activation of astrocytic GABAB receptors triggered intracellular Ca2+ elevations in hippocampal CA1 astrocytes, which led to synaptic potentiation via glutamate release and modulation of presynaptic group I mGluRs (Perea et al. 2016). Astrocytic Ca2+ signaling also plays a role in structural integrity of synapses. It was reported that the IP3 “sponge” mouse model, whose astrocytic IP3 signaling is impaired, has a reduced astrocytic coverage of asymmetric synapses leading to modulation of glutamatergic transmission (Tanaka et al. 2013). In accordance, activity-related structural remodeling of astrocytic processes in the vicinity of synapses was
32 shown to depend on presynaptic activity and to require G-protein-mediated Ca2+ elevations in astrocytes, both in the hippocampus and in the cortex (Perez-Alvarez et al. 2014). Finally, Navarrete et al. (2013) studied cortical and hippocampal human brain samples and showed that increases in intracellular Ca2+ levels of both cortical and hippocampal astrocytes are accompanied by an increase in SICs frequency. These observations confirm that, at least partially, astrocyte Ca2+ elevations are also required to modulate synapses in the human brain. Most of the studies mentioned above regarding synaptic modulation reported consequences of somatic (or in major processes) Ca2+ elevations. Recently, several elegant studies suggested that Ca2+ signals occurring only at astrocyte microdomains are sufficient to modulate synaptic events. Specifically, Ca2+ signals in astrocyte processes of the hippocampal DG are relevant for basal synaptic function, since the application of a Ca2+ chelator or an antagonist for a GPCR predominantly expressed in astrocyte processes decreased synaptic efficacy (Di Castro et al. 2011). Similarly, mGluR-dependent astrocytic Ca2+ signaling was shown to mediate the regulation of basal transmission in CA1 pyramidal neuron synapses (Panatier et al. 2011). The Ca2+ signaling that occurs at the distant processes appears to originate from distinct sources, other than the ER. For instance, TRPA1 channels mediate a transmembrane Ca2+ flux pathway with contributions to basal Ca2+ levels and regulation of interneuron inhibitory synapse efficacy via GABA transporter type 3 (Shigetomi et al. 2011). In addition, the same group showed that TRPA1 activation contributes to the release of D-serine into the extracellular space, which consequently influences NMDAdependent hippocampal plasticity (Shigetomi et al. 2013). Most of the available data resulted from experimental approaches that block general astrocyte Ca2+ elevations, namely by using Ca2+ chelators or genetic IP3R2 deletion. This may explain the current bias toward effects of global (soma and main processes) Ca2+ signals rather than that of focal events (restricted to cellular microdomains). Nevertheless, it is now clear that global and focal signals coexist in specific spatio-temporal maps. Despite the 10to 100-fold scale that distinguishes global and focal signals, both have been shown to precede transmitter release and modulate different forms of synaptic transmission [for review, Volterra et al. (2014)]. Future studies are needed to further discriminate their functional consequences. Altogether, these sets of data suggest that astrocytes integrate brain circuits following two physiological mechanisms (Figure 1.4). First, the activation of astrocytes appears to represent a novel integration mechanism. Astrocyte Ca2+ elevations not only occur upon activation through various excitatory transmitters (e.g. glutamate, ATP, or acetylcholine), but can also be triggered by the inhibitory transmitter GABA. This means that, independently of the nature of the transmitter, astrocytes will be excited/activated
33 (Figure 1.4, left), suggesting they integrate some brain circuits as a redundant layer that reads excitatory and inhibitory neuronal inputs similarly. This concept may add up to the mechanisms of coincidence detection proposed a decade ago by Perea and Araque (Perea and Araque 2005; Perea and Araque 2007). Second, upon activation and Ca2+ elevation, astrocytes release gliotransmitters that ultimately cause either neuronal excitation (e.g. glutamate or D-serine) or inhibition (e.g. ATP degraded to adenosine). Curiously, the available literature indicates that astrocytes provide, in most cases, an excitatory output. Nevertheless, astrocytes also release ATP, which is readily degraded to adenosine and in turn activates A1 receptor leading to synaptic inhibition (Serrano et al. 2006; Martin-Fernandez et al. 2017). This suggests that the type of modulation performed by astrocytes relies not only on the type of transmitter released (which is an intrinsic specificity of the astrocyte), but also on the nature of the receptors expressed by the neighboring cells (which depends on the nature of the neural circuit). In accordance, Martin-Fernandez et al. (2017) also showed that astrocyte-derived ATP results ultimately in excitation or inhibition, depending on the type of receptors found by adenosine. Whether brain circuits are endowed with exclusively excitatory and/or exclusively inhibitory astrocytes, or if these cells play both roles simultaneously (i.e., express machinery to produce, load, and release different transmitters), is still unknown. The first case would not be surprising, since it is the reality for the different neuronal cells (e.g. excitatory glutamatergic vs. inhibitory GABAergic). While recent technical advances have provided fruitful reports of unexpected physiological processes, future studies are needed to confirm their relevance for circuit function. Astrocyte Calcium and Neural Circuits The difficulty to measure and/or manipulate intracellular astrocytic Ca2+ in the intact brain may explain the lack of studies that report on its functional consequences for circuit and behavior computation. Still, recent reports indicate that the Ca2+-dependent modulation of single synapses (as reviewed above) also have an expected impact on brain circuits. Regarding the control of cortical synchronization, it was observed that astrocyte Ca2+ elevations in vivo regulate extracellular glutamate levels, which consequently triggers a slow neuronal rhythm in the brain. This event is characterized by synchronized neuronal firing across different behavioral states, and points to an important regulatory role of astrocytes in cortical circuits (Poskanzer and Yuste 2016). This observation is in line with previous studies performed in brain slices of the cortex and hippocampus that showed reduced neural synchrony upon disruption of astrocytic Ca2+ elevations (Poskanzer and Yuste 2011; Sasaki et al. 2014). Neural rhythmicity in these regions is essential for several behavioral functions
34 such as attention, learning, memory, and control of sleep/wake cycles [for review, Oliveira et al. (2015)]. IP3R2-dependent signaling appears to support ripple-type events that occur during non-theta periods in the CA1 of rodents, a mechanism that might be related to the emotional consequences of social isolation (Tanaka et al. 2017). In the rat trigeminal sensorimotor circuit for mastication, neural rhythmicity arises after sensory encoding. In this circuit, astrocytes actively respond to sensory stimuli by elevating their intracellular Ca2+ levels and, thus, regulating neuronal rhythmic activity, in brainstem slices (Morquette et al. 2015). Furthermore, in thalamo-cortical circuits, astrocytic Ca2+ elevations contributed to the modulation of sensory transmission. More specifically, the activation of mGluR2 triggers intracellular Ca2+ elevations in astrocytes, which are blocked by the astrocyte-specific toxin fluorocitrate. This effect described in vivo and in brain slices is linked to sensory inhibition in the rodent thalamus (Copeland et al. 2017). Furthermore, simultaneous stimulation of whiskers and the nucleus basalis of Meynert revealed that astrocytic Ca2+ elevations precede muscarinic acetylcholine receptor-dependent plasticity in the somatosensory cortex in vivo . This process is dependent on IP3R2 signaling and extracellular D-serine (Takata et al. 2011). In the visual cortex, IP3R2mediated astrocytic Ca2+ elevations are also critical for the integration of visual sensory inputs with the nucleus basalis afferent information (Chen et al. 2012). Moreover, transcranial direct current stimulation (tDCS) was shown to enhance sensory-evoked cortical responses, through elevation of astrocytic intracellular Ca2+ via IP3R2 (Monai et al. 2016). Thus, the authors propose that astrocytic Ca2+ elevations might mediate, at least in part, the recognized improvements obtained by tDCS in neuropsychiatric and neurological conditions. Finally, astrocyte Ca2+ elevation appears to also be involved in the maintenance of homeostatic mechanisms. Astrocytes in brainstem chemoreceptor areas respond to physiological decreases in pH with vigorous elevations in intracellular Ca2+. These were shown to trigger the release of ATP, inducing adaptive increases in breathing (Gourine et al. 2010). Astrocyte Ca2+ elevations appear to control the function of brain circuits at least in two different forms. On one hand, astrocytes appear to control plasticity in synapses occurring between neurons projecting to long distances, similarly to the control of local synapses within one brain region, as discussed above. On the other hand, astrocytes of a specific region appear to support local neural synchronization states, controlling the circuit output, most likely in a multi-synaptic process. Whether the two forms of modulation use similar mechanisms, i.e., whether regional integration is a product of 10 to 1000 of synapses being similarly integrated, is still unknown. However, it seems that both temporal and spatial properties of Ca2+ signals are important for this modulation to occur.
35 Astrocyte Calcium Effects on Behavior The study of rodent models that display altered astrocyte function indicated that astrocytes play important roles in the production of behavior outputs in different dimensions (cognition, emotion, motor, and sensory processing) [(for review, Oliveira et al. (2015)]. As reviewed above, astrocytic Ca2+ signaling is reported to mediate synaptic modulation in different brain circuits. Genetic interference in this astrocytic hallmark has been the main strategy used to assess the role played by astrocyte Ca2+ elevations in behavior, with different laboratories using mostly two approaches: (1) triggering intracellular Ca2+ elevations via chemogenetic activation of GPCR signaling, and (2) inhibiting IP3 signaling by deletion of IP3 receptors or by buffering IP3. The genetic deletion of IP3R2, specifically in about 80% of GFAP-positive cells (at least in the cortex, hippocampus, and substantia nigra), does not appear to influence spatial memory (Petravicz et al. 2014). However, the attenuation of IP3 signaling in GLT1-positive cells led to partial cognitive impairment (Tanaka et al. 2013). Despite the extensive evidence showing a clear influence of astrocyte Ca2+ events on synapses and circuits involved in cognitive behavior (e.g. hippocampus), these two studies provided only modest evidence to support it. We believe that further studies using more specific tools to modulate astrocyte Ca2+ will reveal additional links to cognition. Indeed, we recently showed that spatial learning and memory rely on astrocyte exocytosis, which is a Ca2+-dependent mechanism, and therefore might be dependent on astrocyte integration of surrounding activity (Sardinha et al. 2017). IP3-dependent astrocytic Ca2+ signaling does not seem to be related to anxiety-like behavior as indicated by studies using a constitutive (Cao et al. 2013; Tanaka et al. 2013) or conditional (Petravicz et al. 2014) IP3R2 deletion. Regarding the depressive-like behavior, the available data is not consistent. While the constitutive deletion of IP3R2 was shown to trigger some forms of a depressive phenotype (Cao et al. 2013), these are completely absent in the model with conditional deletion (Petravicz et al. 2014). Curiously, the GFAP-MrgA1 model that allows stimulation of astrocyte Ca2+ displayed decreased learned helplessness in the forced swim test (Cao et al. 2013). Further experimentation is required to clarify these apparently contradictory results. Finally, a recent study indicated that the astrocytic control of excitatory/inhibitory inputs to the central amygdala is crucial for fear-related behavior (Martin-Fernandez et al. 2017). Regarding motor function, the activation of IP3 “sponge” or deletion of IP3R2 does not seem to interfere with exploratory behavior (Cao et al. 2013; Tanaka et al. 2013; Petravicz et al. 2014). Interestingly, the chemogenetic activation of astrocyte Gq-coupled signaling led to an impairment in motor coordination in an IP3R2-independent manner (Agulhon et al. 2013). The specific ablation of IP3R2 in GLAST-positive cells
36 resulted in an impairment in motor-skill learning of a forelimb reaching task (Padmashri et al. 2015). More so, the attenuation of IP3/Ca2+ signaling in astrocytes resulted in modulation of the rodent sleep, by increasing the time spent and frequency periods in rapid eye movement phase (Foley et al. 2017). While the research carried out so far has provided us with critical knowledge of the influence of astrocytic Ca2+ in synaptic or circuit function, further research should be performed to complement the rather sparse evidence on the influence of astrocytes on behavior. Moreover, further evidence will help clarify the discrepant results and reconcile the existent observations. For instance, the different observed cognitive outcomes might be related to the different mechanism used to target intracellular Ca2+ (full deletion of IP3R2 vs. IP3 “sponge”), or by the promotor used to drive astrocyte specificity (GFAP vs. GLT-1) that could result in differential regional expression. Moreover, there are at least five different IP3R2 KO strains with different genetic backgrounds (Oliveira et al. 2015); this might not be critically relevant for studies of synaptic plasticity in brain slices, but it may easily lead to distinct observations in behavior tests. Additionally, the different studies should be conducted in a standardized form to allow the linear comparison of the results. All rodent models reported above result on drastic modulation of global Ca2+ elevations. Intracellular Ca2+ rises may trigger a multitude of signaling pathways. Thus, it must always be considered that the interference with global Ca2+ events may trigger several simultaneous consequences, which might lead to confounding results. Although these models might be very useful when carefully assessed, the research field requires the development of more specific models to allow temporal and, more importantly, local control of Ca2+ signals in physiological processes. The IP3R2 KO mouse model Due to the high expression of IP3R2 specifically in astrocytes, the IP3R2 knockout (IP3R2 KO) mouse model naturally emerged as a tool to study astrocytic Ca2+ signaling (Li et al. 2005; Petravicz et al. 2008). These mice lack global Ca2+ elevations, which correspond to the slower, long-lasting events that are mainly confined to the soma and main processes. However, they still display IP3R2-independent Ca2+ signals in astrocytic microdomains. These signals derive from alternative Ca2+ sources as the plasmalemmal Ca2+ influx or the mitochondria (Srinivasan et al. 2015; Rungta et al. 2016; Agarwal et al. 2017). The IP3R2 KO mouse model was generated at the Ju Chen laboratory (Li et al. 2005). It had a strong Black Swiss background, that could cause retinal degeneration and blindness with a negative impact on behavioral performance (Clapcote et al. 2005). Our lab and others circumvented this constraint by backcrossing the founder line to C57BL/6J mice (Wang et al. 2012b; Chen et al. 2012; Petravicz et al. 2014; Feriod et al.
37 2014; Biesecker et al. 2016; Guerra-Gomes et al. 2018; Pinto-Duarte et al. 2019). The variety of strains derived from the initial models generated in Prof. Ju Chen (Li et al. 2005) and Prof. Katsuhiko Mikoshiba (Futatsugi et al. 2005) are summarized in Table 1.2. For the sake of simplicity, the mouse line generated by Li et al. (2005) is referred as IP3R2 KO, and whenever crossed with C57BL/6J wild-type mice as IP3R2 KO (C57BL/6). The mouse line generated by Futatsugi et al. (2005) has this reference added to the IP3R2 KO designation and crossings with other mouse lines are specifically mentioned. The IP3R2 KO mice generated by Li et al. (2005) are viable and fertile and display no overt behavioral abnormalities (Petravicz et al. 2008). Due to the importance of IP3R2 in hepatocytes, Feriod and coauthors (2014) dissected the role of this receptor in metabolic functionality. They observed that C57BL/6 IP3R2 KO mice had a normal energy balance, glucose production and tolerance, and insulin sensitivity. Moreover, no alterations in liver morphology were found. Despite the recognized importance of the IP3 signaling for astrocytic Ca2+ elevations, contrasting results were reported regarding synaptic transmission and plasticity in these mice. For instance, cholinergic-induced presynaptic LTP in cortical (Takata et al. 2011) and hippocampal (Navarrete et al. 2012) regions was shown to be affected, while induction of NMDAR-dependent LTP was not (Agulhon et al. 2010). Recently, two studies demonstrated that IP3R2 KO mice present an impaired NMDAR-dependent hippocampal long-term depression (LTD) (Pinto-Duarte et al. 2019; Navarrete et al. 2019). Depending on the IP3R2 KO model used, several laboratories reported distinct observations for different behavioral dimensions. The outcomes obtained from studies using IP3R2 KO mouse model or IP3R2 KO crossed with other models are summarized in Table 1.2. Several studies have demonstrated that interfering with IP3-dependent Ca2+ signaling affects synaptic, circuit and behavioral functioning, while others report no changes related to these brain functions (Guerra-Gomes et al. 2017). Moreover, some evidences point to a role of IP3R2-dependent Ca2+ in K+ homeostasis and in pathological conditions (e.g. stroke, injury, neurodegenerative disorders) (Wang et al. 2012a; Dong et al. 2013a; Kanemaru et al. 2013; Li et al. 2015; Kim et al. 2016; Rakers and Petzold 2017; Heuser et al. 2018). Interestingly, an overall role of neuroprotection against an insult or a pathological state was observed in these studies. Contrarily, the control of neurovascular coupling was shown to be independent of an IP3R2-Ca2+ mechanism (Nizar et al. 2013; Bonder and McCarthy 2014). Novel tools are needed to draw further conclusions regarding the functional influence of astrocytic Ca2+ signals. Recently, Yu et al. (2018) developed a mouse model that constitutively extrude cytosolic Ca2+ in astrocytes, targeting specific spatiotemporal aspects of astrocyte Ca2+ signaling in the striatum. This approach revealed an important role for astrocytic Ca2+ in the control of a striatum-dependent behavior.
38 Thus, the development of novel genetic tools with higher spatial and temporal resolution will be pivotal to dissect the role of astrocytic Ca2+ signaling in brain functioning. Table 1.2 - Functional consequences of the interference with IP3R2 expression Symbol legend: ↑, increased; ↓, decreased; =, no change. Model Alteration/modulation Approach Phenotype References BEHAVIOR COGNITION GFAP-IP3R2 cKO (C57BL/6) Lack of intracellular global Ca2+ signaling in GFAP+ astrocytes; C57BL/6 background; Spatial Reference Memory and Reversal Learning (MWM) = long-term spatial memory = reversal learning (Petravicz et al. 2014) GLT1-IP3 “sponge” Attenuation of intracellular global Ca2+ signaling in GLT-1+ astrocytes Spatial Reference Memory (MWM) and Fear Conditioning ↓ long-term memory (Tanaka et al. 2013) Appps1+/− x IP3R2 KO Lack of intracellular global Ca2+ signaling in astrocytes in a mouse model of Alzheimer’s disease (AD) Spatial Reference Memory (MWM) ↑ spatial memory and network hyperactivity in a mouse model of AD (Reichenbach et al. 2018) IP3R2 KO (C57BL/6) Lack of intracellular global Ca2+ signaling in astrocytes Y-maze Two-Trial Place Recognition (2TPR) = cognitive performance (PFCdependent) with aging (GuerraGomes et al. 2018) IP3R2 KO (C57BL/6) Lack of intracellular global Ca2+ signaling in astrocytes Y-maze (spontaneous alternations) ↓ spontaneous alternations (working memory) (Pinto-Duarte et al. 2019) Barnes maze test ↓ remote spatial memory = recent spatial memory Cued and contextual fear conditioning ↓ remote fear memory = recent fear memory Novel object recognition ↓ long-term recognition memory EMOTION Anxious-like behavior IP3R2 KO Lack of intracellular global Ca2+ signaling in astrocytes Elevated Plus Maze = anxious-like behavior ↓ ATP levels (Cao et al. 2013) GLT1-IP3 “sponge” Attenuation of intracellular global Ca2+ signaling in GLT-1+ astrocytes Open Field = anxious-like behavior (Tanaka et al. 2013) GFAP-IP3R2 cKO (C57BL/6) Lack of intracellular global Ca2+ signaling in GFAP+ astrocytes Elevated Plus Maze Activity Box = anxious-like behavior (Petravicz et al. 2014) IP3R2 KO (C57BL/6) Lack of intracellular global Ca2+ signaling in astrocytes Light/dark box = anxious-like behavior (Pinto-Duarte et al. 2019)
39 Depressive-like behavior IP3R2 KO Lack of intracellular global Ca2+ signaling in astrocytes Sucrose Preference Test ↑ anhedonia (Cao et al. 2013) Forced Swim Test ↑ learned helplessness (Cao et al. 2013) Coat Score ↑ depressive behavior (Cao et al. 2013) IP3R1, IP3R2 and IP3R3 knockdown Mice carrying a cerebral knockdown for each IP3R isoform Forced Swim Test ↓ depressive behavior (Galeotti et al. 2008) GFAP-IP3R2 cKO (C57BL/6) Lack of intracellular global Ca2+ signaling in GFAP+ astrocytes Tail Suspension = depressive-like behavior (Petravicz et al. 2014) Impulsivity GFAP-IP3R2 cKO (C57BL/6) Lack of intracellular global Ca2+ signaling in GFAP+ astrocytes Acoustic startle = sensory-motor gating and prepulse inhibition (Petravicz et al. 2014) MOTOR Motor activity and coordination IP3R2 KO Lack of intracellular global Ca2+ signaling in astrocytes Open Field = exploratory/locomotor behavior (Cao et al. 2013) GLT1-IP3 “sponge” Attenuation of intracellular global Ca2+ signaling in GLT-1+ astrocytes Open Field = exploratory behavior (Tanaka et al. 2013) GFAPhM3Dq x IP3R2 KO Chemogenetic control of astrocytic Gqcoupled signaling independent of IP3 Open Field = exploratory behavior (Agulhon et al. 2013) Rotarod ↓ motor coordination Righting reflex analysis ↓ motor coordination IP3R2 KO (C57BL/6) Lack of intracellular global Ca2+ signaling in astrocytes Electrocorticogram ↓ astrocyte-cortical upstate synchrony (Wang et al. 2012b) GLASTCreIP3R2fl/fl Lack of intracellular global Ca2+ signaling in GLAST+ astrocytes Forelimb Reaching Task ↓ motor-skill learning (Padmashri et al. 2015) GFAP-IP3R2 cKO (C57BL/6) Lack of intracellular global Ca2+ signaling in GFAP+ astrocytes Open Field Rotarod = motor function and exploratory behavior = motor coordination and learning (Petravicz et al. 2014) IP3R2 KO (C57BL/6) (Futatsugi et al. 2005) Lack of intracellular global Ca2+ signaling in astrocytes Rotarod = motor function (Kim et al. 2016) IP3R2 KO (C57BL/6) Lack of intracellular global Ca2+ signaling in astrocytes Locomotor activity test = motor function (PintoDuarte et al. 2019)
46 instance, the hippocampus has a recognized role in the spatial and temporal distinction of different stimuli (e.g. textures or shapes) and is implicated in avoidance tests. It is further described as necessary to suppress an originally formed memory, and for memory recall over 3-4 weeks after acquisition. Another limbic region involved in cognitive processing is the amygdala. This brain region is important in learning and memory processes usually associated with an aversive stimulus (e.g. footshock). Due to its close implication in motor functions, the cerebellum is required for motor learning, whereas the striatum is critical for motor and reward system functioning. The rhinal cortex, a cortical region of the medial temporal lobe, has been implicated in object recognition and the cortical regions that receive direct outputs from the ventral hippocampus – the PFC – are fundamental for working memory and memory extinction tasks (Sousa et al. 2006). An important concern when working with full knockout models is the relevance of the deleted gene during developmental stages. Genetic manipulation can also lead to early developmental alterations that could influence adult cognitive behavior. Thus, the achievement of milestones through an examination of somatic parameters and neurological reflexes should be addressed to evaluate the developmental progression (Hill et al. 2008). Additionally, this analysis can be complemented by assessing general phenotypical behavior in adult mice through the assessment of motor behavior and exploratory drive. Cognitive tests encompass interacting aspects of learning and memory, so by testing several paradigms we will be able to obtain a comprehensive readout of cognitive function. Learning and memory functions can be assessed through different mazes and paradigms. In this thesis, we tested several PFC and/or hippocampal-dependent cognitive tasks. Each of the cognitive tests used is depicted in Figure 1.6.
47 The Morris Water Maze (MWM) is a gold-standard test to evaluate spatial navigation in an open swimming area surrounded by visual cues. In this test, several paradigms can be assessed: RM, WM and reversal learning. All of these involve the use of visual cues to successfully reach a hidden platform that can maintain (RM) or shift (WM) its location during several trials (Vorhees and Williams 2006). Cognitive tests are designed with the premise that animals will have willingness to participate, either because they are exposed to a positive (sugary pellet vs regular food) or negative (footshock) reinforcer. A great example of a learning test associated with a negative reinforcer is the Contextual Fear Conditioning (CFC). In this test, animals associate an aversive electric shock with a cue (light or tone) and a specific context (Gu et al. 2012). In contrast, both the Y-maze task and the Novel Object Recognition (NOR) test are based on the natural drive of rodents to explore novelty (Leger et al. 2013; Holter et al. 2015) Astrocytic modulation of cognitive behavior Astrocytic signaling plays a key role in the control of brain homeostasis and communication. Growing evidence point to astrocytes as key regulators of cognitive processing (Gibbs et al. 2008; Pannasch and Rouach 2013; Fields et al. 2014; Oliveira et al. 2015; Dallerac and Rouach 2016; Santello et al. 2019). Over the past years, pharmacological and genetic tools were developed to study astrocytic function, Figure 1.6 - Illustrative scheme of the behavioral tests used for cognitive assessment. The scheme represents the behavioral tests used to assess cognitive function in this thesis. Morris Water Maze (MWM) evaluates spatial recognition memory, as the animal needs to find a hidden platform in a pool guided by spatial cues. Contextual Fear Conditioning (CFC) allows the assessment of fear memory acquisition since animals are exposed to an aversive stimulus (light followed by a shock). Y-maze 2-Trial Place Recognition (2TPR) takes advantage of the ability of rodents to explore novelty, to evaluate their memory for novel locations. Novel object recognition task (NOR) evaluates the ability of rodents to memorize a familiar (previously explored) and novel object in an arena.
48 targeting specific features such as: i) astrocytes' ability to “sense”; ii) elevate its intracellular Ca2+ levels in response to neuronal activity and iii) respond to the surrounding activity by releasing gliotransmitters. These tools allowed a better comprehension of the role of astrocytes in brain functioning, by tackling specific astrocytic functions and exploring its implications, for instance, in cognitive performance. For example, astrocytes express a variety of functional receptors in their membranes. Robin and co-authors (2018) showed that genetic deletion of CB1R specifically in astrocytes impairs hippocampal LTP and longterm recognition memory. These observations were rescued by application of the gliotransmitter, D-serine. Similarly, we previously demonstrated that conditional blockade of exocytosis in astrocytes, using the dnSNARE mouse model, critically impairs cognitive function. These deficits were completely restored after D-serine application (Sardinha et al. 2017). A study using another approach to interfere with astrocytic exocytosis, through the expression of tetanus toxin, revealed that these animals presented an impairment in the performance of a recognition memory task (Lee et al. 2014). Contrastingly, modulation of astrocytic function can also result in cognitive enhancement. This was elegantly shown by Mederos et al. (2019), using the specific expression of a G-protein-coupled photopigment to trigger astrocytic Ca2+ elevations. This approach enhanced episodic-like memory in an object recognition task. In the same way, the expression of a modified form of the human M3 muscarinic receptor in CA1 astrocytes was able to enhance fear memory, in the CFC paradigm and spatial recognition memory in the T-maze (Adamsky et al. 2018). The role of astrocytes in the maintenance of neuronal networks has also been reported to change in response to a pathological state. In such cases, neuron-astrocyte interaction undergoes functional, morphological and molecular changes that strongly affect cognitive function. For example, Alzheimer’s disease (AD) is a known neurodegenerative disorder characterized by marked cognitive deficits. Orr and colleagues (2015) found that an AD mouse model presented increased expression of astrocytic A2A receptors that contributed to the observed cognitive impairments. The genetic deletion of these receptors in astrocytes was able to enhance memory performance in the MWM test. Moreover, the lack of IP3dependent Ca2+ signaling in astrocytes in a mouse model of AD enabled the restoration of both spatial memory and network hyperactivity (Reichenbach et al. 2018). Furthermore, in the experimental autoimmune encephalomyelitis model of multiple sclerosis, the astrocytic TNF receptor type 1 signaling was shown to be implicated in the impaired fear memory observed (Habbas et al. 2015). Accordingly, several brain diseases lead to impaired cognitive processing. An aberrant astrocytic activity has been demonstrated to influence these outputs and even accelerate the undesirable events that occur through a pathological state.
49 Astrocytes in disease: involvement in depression Astrocytes participate in the control of several functions related to CNS normal development and functioning. Upon injury or under pathological conditions, astrocytes undergo a phenotypical alteration, changing their morphology and molecular profile, which can be protective or detrimental to the diseased tissue (Khakh and Sofroniew 2015). Generally, it is considered that after being exposed to an insult, astrocytes become “reactive” and respond through a process called astrogliosis (Sofroniew 2014). Further analysis into the molecular profile of reactive astrocytes revealed an A1 (neurotoxic) or A2 (neuroprotective) astrocytic phenotype (Zamanian et al. 2012; Liddelow et al. 2017). A growing body of evidence reported an altered astrocytic Ca2+ signaling in response to brain diseases, such as neurological and psychiatric disorders (Nedergaard et al. 2010; Shigetomi et al. 2016). However, the mechanisms underlying the development of the deleterious effects after a pathological insult are still unexplored. In this thesis, after observing a positive modulation of hippocampal-dependent behavior, we decided to test its modulatory potential in a model of depression, which is a disease that affects cortico-limbic circuits with impact both in cognitive and emotional behavior. Depression is a neurobiological, multidimensional disorder and the leading cause of disability worldwide (WHO 2017). This is a highly heterogeneous disorder, with a variable course and an inconsistent response to treatment. Depressive patients present alterations in distinct behavioral domains such as mood, anxiety, and cognition (Bessa et al. 2009b). The unpredictable chronic mild stress (uCMS) protocol developed by Willner et al. (1987) induces a depressive-like phenotype in mice and rats, due to the exposure to several stressors in a random and unpredictable fashion. This protocol is highly translatable to the human disease since it has a multi-dimensional behavioral impact (Bessa et al. 2009b). In fact, studies performed in our lab using the uCMS animal model of depression have reported alterations in mood, anxiety, and cognition. Moreover, rats exposure to uCMS leads to alterations in the hypothalamic– pituitary–adrenal (HPA) axis activation, neural plasticity, cytogenesis and different molecular profiles (Bessa et al. 2009b; Mateus-Pinheiro et al. 2013; Patricio et al. 2015; Alves et al. 2017). The hippocampal formation is affected by chronic stress, which paved the way to look to other interconnected brain regions, such as the PFC and the amygdala (McEwen et al. 2016). In fact, our lab has previously shown a clear neuronal compromise in the PFC (Dias-Ferreira et al. 2009) and the hippocampus (Mateus-Pinheiro et al. 2013) after exposure to chronic stress. In addition, imaging studies performed in our lab support these observations by reporting structural atrophy in cortico-limbic regions and others, after stress exposure (Magalhaes et al. 2018).
50 The recognized role of astrocytes in brain function led to the disclosure of its importance in mood disorders. A reduced immunoreactivity of specific astrocytic markers (GFAP and GS) was observed in cortico-limbic regions in rodent models of depression (Musholt et al. 2009; Gosselin et al. 2009) and human patients (Miguel-Hidalgo et al. 2000; Cotter et al. 2002). Furthermore, Golgi-Cox staining analysis revealed an altered astrocytic cortical morphology in samples from depressed suicides (Torres-Platas et al. 2011). Our group and others demonstrated that, in rats, the specific ablation of astrocytes in the PFC leads to cognitive deficits in a WM task (Lima et al. 2014) and induce an anhedonic phenotype (Banasr and Duman 2008; Banasr et al. 2010), two major hallmarks of depression. Additionally, glutamatergic excitotoxicity has been described as an important feature in the installation of the disease, due to the defective neurotransmitter uptake by astrocytes (Rajkowska and Stockmeier 2013). In line with this, the administration of a drug that induces the overexpression of the astrocyte glutamate transporter GLT-1 – ceftriaxone – was shown to be able to exert an anti-depressant effect (Mineur et al. 2007). Contrarily, interfering with glutamate uptake namely by blocking GLT-1, leads to a depressive-like phenotype (Lee et al. 2007; John et al. 2012) and impairs spatial memory (Bechtholt-Gompf et al. 2010). Despite the emerging importance of astrocytes in the context of brain disorders, it is still unexplored the functional mechanisms underlying its contribution to the pathophysiology of depression. So far, the role of signaling dependent on global astrocyte Ca2+ in the context of mood disorders, namely depression is poorly understood. Astrocytes in aging Aging is a lifelong process characterized by an accumulation of several detrimental changes that lead to an increased risk of disease and death (Tosato et al. 2007). In the brain, aging alters molecules, cells, vasculature, brain size and cognition (Peters 2006). Imaging studies revealed that aging is accompanied by alterations in regional brain volumes in areas related to cognitive function, motor behavior and emotional processing (Amorim et al. 2018). Specifically, the PFC and basal ganglia show a particular vulnerability to the aging process, translated into functional and morphological (atrophic) alterations (Burke and Barnes 2006; Berchtold and Cotman 2009). Based on this, heathy aging affects cognitive functions mainly related with memory, executive function, and processing speed. Contrastingly, in pathological aging, a global cognitive dysfunction is observed (Berchtold and Cotman 2009). In healthy aging, several brain cell types become compromised and contribute to cognitive impairments. Several studies tried to dissect the underlying morphological and molecular alterations occurring in the aging body. However, conflicting results lead to a myriad of controversial observations. Regarding neuronal cell
51 number, some reports suggested a decrease in the dorsal PFC (Stranahan et al. 2012), mPFC and primary visual cortex of rodents (Yates et al. 2008), while others revealed no changes in neuronal counts in neocortical regions of human aged brains (Freeman et al. 2008). Similarly, dendritic morphology and spine analysis revealed a huge variability of results, from no changes to an increased dendritic complexity upon aging (Konsolaki and Skaliora 2015). The lack of a clear neuronal influence that could explain the behavioral results observed with aging, led some authors to explore the role of other cells involved in synaptic control such as astrocytes. Aging triggers functional, morphological and molecular alterations in astrocytes (Palmer and Ousman 2018). Age-related changes alter astrocytic ability to properly maintain a homeostatic brain environment. A reduction in astrocytic - synaptic coverage, neurovascular and metabolic support has been observed and their consequences in the aged hippocampus were reviewed by Ojo et al. (2015). In addition, age-dependent oxidative stress in middle-aged mice was shown to lead to an increase in reactive oxygen species and Ca2+ overload in astrocytes (Ishii et al. 2017). Recently, Gomez-Gonzalo et al. (2017) demonstrated that spontaneous and neuronal-activity evoked astrocytic Ca2+ signals remain unaltered with normal aging, using hippocampal and cortical slices. Moreover, authors showed that in the context of AD, interfering with IP3R2-dependent Ca2+ signaling in astrocytes contributes to the disease installation. Contrastingly, neocortical astrocytes from WT and dnSNARE mice displayed an age-related decrease in spontaneous Ca2+ elevations. This alteration was counterbalanced by exposing animals to an environmental enrichment or caloric restriction approach (Lalo et al. 2018). Astrocytes are recognized to undergo alterations in their main structure with aging. Typically a process called astrogliosis occurs and a general hypertrophy is observed (Cotrina and Nedergaard 2002). A comprehensive analysis studied GFAP-immunostaining in three brain regions (entorhinal cortex, DG and CA1 hippocampus) from young and aged mice (Rodriguez et al. 2014). These authors strikingly showed an increase in GFAP-stained astrocytic structure in DG and CA1 regions, contrasting with a reduction in this staining at the entorhinal cortex. These results point to a heterogeneous response of astrocytes to age-dependent changes. The breakthrough of transcriptomic analysis revealed that aging alters astrocytic gene expression patterns in humans (Soreq et al. 2017) and mice (Boisvert et al. 2018). Astrocytes specifically isolated from aged mice display an upregulation of genes related with synapse elimination and immune pathways, a decrease in the expression of genes associated with cholesterol production, and major gene alterations in brain areas intimately related to the aging process, as the cerebellum and the hypothalamus (Boisvert et al. 2018). Clarke et al. (2018) added that, in healthy aging, astrocytes take on an A1 phenotype and have an increased number of differentially expressed genes, particularly in the hippocampus and striatum.
52 Together, compelling evidences point to an important role of astrocytes in brain function control in the context of depression and aging. Thus, it urges to explore the potential of global astrocyte Ca2+ signaling to modulate aging-related molecular, structural and cognitive deficits. Aims of this thesis Astrocytes are acknowledged for their role in brain function, mainly due to the identification of their intracellular Ca2+ dynamics and the development of imaging techniques. The view of these cells as simple passive, non-excitable cells, supporting neuronal activity is now outdated. About 30 years ago, the discovery of astrocyte intracellular Ca2+ elevations completely revolutionized the field (Cornell-Bell et al. 1990; Charles et al. 1991). These elevations represent a proxy of astrocyte activity and prompted neuroscientists to study its role in physiological and pathological conditions. Astrocytic Ca2+ elevations were found to emerge from different sources and display complex spatio-temporal properties (Bazargani and Attwell 2016). In the brain, astrocytes are an organized network that respond to activation via Ca2+ signaling, driving orchestrated responses. Therefore, astrocytic Ca2+ impacts synapses, circuits and behavior (Guerra-Gomes et al. 2017). Global astrocytic Ca2+ elevations occur mainly through a Gq-GPCR signaling pathway that leads to the production of IP3, following the activation of IP3R2, which triggers the release of Ca2+ from the ER. These global Ca2+ signals were shown to modulate different forms of synaptic transmission in several brain regions, with a potential impact to modulate circuits and behavior. The main goal of this thesis was to investigate the role of global astrocytic Ca2+ signaling in cortico-limbic function and behavior. For that, we took advantage of the IP3R2 KO mouse model that lacks global Ca2+ signaling in astrocytes. We performed a longitudinal characterization of this mouse model by assessing its impact during postnatal development, adulthood (healthy and in a depression context) and aging, at behavioral, structural and molecular levels. We were interested in understanding the impact of astrocytic Ca2+ signaling in cognitive function, mainly dependent on the hippocampus and the PFC. To address this main goal we aimed at: i) Characterize the impact of IP3R2 deletion during postnatal development (Chapter 2); ii) Evaluate the role of IP3R2-dependent astrocytic Ca2+ signaling in cognitive function by performing a detailed behavioral, structural and molecular characterization (Chapter 3); iii) Test a putative involvement of astrocytic IP3R2-dependent Ca2+ signaling in the deleterious effects of uCMS exposure (Chapter 4). iv) Test the protective potential of IP3R2-dependent Ca2+ signaling in astrocytes during aging (Chapter 5);
53 CHAPTER 2 Sónia Guerra-Gomes, Eduardo Loureiro-Campos, Vanessa Morais Sardinha, Diana Sofia Marques Nascimento, João Filipe Viana, Nuno Sousa, Luísa Pinto, João Filipe Oliveira IP3R2 KO mice display a normal somatic and neurological development (Manuscript submitted) (2019)
54 Title: IP3R2 KO mice display a normal somatic and neurological development Authors: Sónia Guerra-Gomes1,2, Eduardo Loureiro-Campos1,2, Vanessa Morais Sardinha1,2, Diana Sofia Marques Nascimento1,2, João Filipe Viana1,2, Nuno Sousa1,2, Luísa Pinto1,2, João Filipe Oliveira1,2,3 Affiliations: 1 - Life and Health Sciences Research Institute (ICVS), School of Medicine, University of Minho, 4710057 Braga, Portugal 2 - ICVS/3B’s - PT Government Associate Laboratory, Braga/Guimarães, Portugal 3 - IPCA-EST-2Ai, Polytechnic Institute of Cávado and Ave, Applied Artificial Intelligence Laboratory, Campus of IPCA, Barcelos, Portugal Corresponding author: João Filipe Pedreira de Oliveira (PhD), Life and Health Sciences Research Institute (ICVS), School of Medicine, University of Minho, Campus de Gualtar, 4710-057 Braga, Portugal, Tel: +351-253-604871, Fax: +351-253-604809, [email protected] ACKNOWLEDGMENTS The authors are grateful to Prof. Alfonso Araque and Prof. Ju Chen for sharing the mice line. They acknowledge also funding from national funds through Foundation for Science and Technology (FCT) fellowships to SGG, ELC, VMS, CEEC to DSMN and IF grants to LP (IF/01079/2014) and JFO (IF/00328/2015); Bial Foundation Grants 207/14 and 037/18 to JO, and 427/14 to LP; Northern Portugal Regional Operational Programme (NORTE 2020), under the Portugal 2020 Partnership Agreement, through the European Regional Development Fund (FEDER) (NORTE-01-0145-FEDER000013); FEDER Funds, through the Competitiveness Factors Operational Programme (COMPETE), and The National Fund, through the FCT (POCI-01-0145-FEDER-007038).
55 Abstract Astrocytes are key players in the regulation of brain development and function. They sense and respond to the surrounding activity by elevating their intracellular calcium (Ca2+) levels. These astrocytic Ca2+ elevations emerge from different sources and display complex spatio-temporal properties. Ca2+ elevations appear spatially distributed in global (soma and main processes) and/or focal regions (microdomains). The inositol 1,4,5-triphosphate receptor type 2 knockout (IP3R2 KO) mouse model lacks global Ca2+ elevations in astrocytes and it has been recently used by different laboratories in the glia field. However, this model is a constitutive knockout posing the natural concern that the absence of IP3R2 during development may trigger compensating phenotypes, which could bias results of experiments using developing or adult mice. Therefore, we performed a detailed neurodevelopmental evaluation of male and female IP3R2 KO mice, during the first 21 days of life. This evaluation consisted of a detailed examination of somatic parameters and neurological reflexes. Our results show that male and female IP3R2 KO mice display a normal growth, as compared with wild-type (WT) mice. IP3R2 KO mice reached all mature neurological reflexes within the time windows of WT mice. As expected, this normal development results in normal locomotion and exploratory drive in this mouse model in adulthood. Overall, our results show that the IP3R2 KO mouse is a reliable model to study the functional impact of global IP3R2-dependent astrocytic Ca2+ elevations. Keywords Astrocytes, calcium signaling, IP3R2, development, behavior
62 Figure 2.1 - Male and female IP3R2 KO mice gain weight similarly to controls and display a normal genital development. (A) Representative scheme of the experimental timeline used, depicting each developmental parameter according to the tested function (grey – somatic parameters, blue – tactile reflex, pink – auditory reflex, orange – motor function, green – labyrinthine reflex, coordination, and strength). (B) Body weight evolution from postnatal day 1 to postnatal day 21 of WT and IP3R2 KO mice (male and female). (C) Anogenital distance measurement from postnatal day 3 to postnatal day 21 of WT and IP3R2 KO mice (male and female). Males are represented in left panels and female mice are depicted in right panels (WT – black, IP3R2 KO – red). Data plotted as mean ± SEM. n=9-10 per group. Postnatal day 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 Somatic Weight Anogenital distance Eye opening Tactile reflex Rooting Ear tw itch Auditory reflex Auditory startle Motor Open Field Walking Labyrinthine reflex, coordination and strength Surface righting Air righting Wire suspension Grasping Postural reflex Negative geotaxis Cliff aversion Anogenital distance (mm) 0 3 6 9 12 15 18 21 0 2 4 6 8 10 Postnatal day Weight (g) A B C Male Female
63 IP3R2 KO mice acquire all neurological milestones normally Neurodevelopmental evaluation included the examination of several neurological parameters related to sensorial (Figure 2.1A, blue and pink) and motor functions (Figure 2.1A, orange), labyrinthine reflex, coordination and strength (Figure 2.1A, green) in male and female WT and IP3R2 KO mice. Tactile reflex development was evaluated in rooting and ear twitch tests (Figure 2.2B). Our results show that male WT and IP3R2 KO mice display a similar mature response in the rooting test ( U = 34.5; p = 0.40), which was equivalently observed for females ( U = 35.5; p = 0.29). More so, ear twitch test revealed no significant differences between WT and IP3R2 KO mice from both genders (male: U = 41.5; p = 0.78; female: Figure 2.2 - Male and female IP3R2 KO mice retain a normal acquisition of somatic and sensory-motor parameters. (A) Postnatal acquisition of a somatic milestone (eye opening, grey background). (B-D) Postnatal acquisition of neurological reflexes related with sensorial (B,C) and motor function (D). (B) Rooting and ear twitch tests were performed to evaluate the developmental acquisition of tactile reflex (blue background). (C) Auditory startle evaluated the acquisition of a mature auditory reflex (pink background). (D) Open Field and Walking are parameters for the assessment of a mature locomotor response (orange background). WT mice are represented in black and IP3R2 KO mice are represented in red (males, left column; females, right column). Data plotted as median. n=9-10 per group. Tactile reflex Motor Rooting Ear twitch Open field Walking Auditory reflex Auditory startle Eye opening Somatic A B C D
64 U = 37.5; p = 0.36). Auditory system development was assessed by the auditory startle test. We observed that male WT and IP3R2 KO newborn mice equally acquired this milestone (Figure 2.2C; U = 38.0; p = 0.58), and females from both genotypes displayed a similar mature response ( U = 32.5; p = 0.19). The achievement of motor abilities is a good indicator of neurological development. We evaluated motor function in the open field traversal and walking tests (Figure 2.2D). Our results show an equal acquisition of the mature response in the open field traversal for male ( U = 42.5; p = 0.86) and female ( U = 39.5; p = 0.44) mice from both genotypes. Furthermore, male WT and IP3R2 KO mice equally acquired a mature walking ability ( U = 38.5; p = 0.63), which was also observed for females from both genotypes ( U = 43.5; p = 0.63). The neurological milestones also comprise several tests to evaluate the vestibular system formation, namely surface righting, air righting, postural reflex, negative geotaxis and cliff aversion (Figure 2.3). In the surface righting test, we evaluated labyrinthine reflex and coordination. We found no differences between male WT and IP3R2 KO mice in this test ( U = 40.5; p = 0.73). A similar mature response was also observed for female WT and IP3R2 KO mice ( U = 33.5; p = 0.21). In the air righting test, male WT and IP3R2 KO displayed an analogous mature response ( U = 34.0; p = 0.38), which was also observed for females ( U = 49.5; p = 0.97). Moreover, we found that genetic deletion of IP3R2 did not alter postural reflex acquisition in males ( U = 37.5; p = 0.56) and females ( U = 38.5; p = 0.38). In the negative geotaxis test the mature response of male IP3R2 KO mice was similar to WT littermates ( U = 42.0; p = 0.83). Female WT and IP3R2 KO mice also equally acquired this developmental milestone ( U = 43.0; p = 0.59). Lastly, cliff aversion test showed that male WT and IP3R2 KO mice present a mature response approximately at the same postnatal day ( U = 42.0; p = 0.82), which was also confirmed for female IP3R2 KO mice ( U = 31.0; p = 0.16). Another important measure of postnatal development is strength. We assessed this parameter by performing wire suspension and grasping test (Figure 2.3). Our results show that male WT and IP3R2 KO display a similar mature response in wire suspension ( U = 44.5; p = 0.98), which was again confirmed for IP3R2 KO female mice ( U = 26.0; p = 0.06). Overall, this detailed assessment of neurological parameters shows that IP3R2 KO mouse model acquires a mature response for sensory-motor functions, vestibular area-dependent tasks, and strength similarly to WT.
65 Figure 2.3 - Male and female IP3R2 KO mice display normal labyrinthine reflex, coordination and strength. Surface righting, air righting, postural reflex, negative geotaxis, cliff aversion, wire suspension and grasping as measures of vestibular system formation and strength acquisition. WT mice are represented in black and IP3R2 KO mice are represented in red (males, left column; females, right column). Data plotted as median. n=9-10 per group. IP3R2 KO adult mice display normal motor and exploratory behavior Developmental milestones showed that impairment of IP3-dependent Ca2+ signaling in astrocytes does not interfere with the acquisition of somatic and neurological parameters. Nevertheless, we also aimed to assess locomotor and exploratory function in adult WT and IP3R2 KO mice, both in males and females. The open field test provided us with measures of ambulatory behavior and motor function, such as distance traveled and number of vertical rearings (Figure 2.4A). Male and female IP3R2 KO mice traveled similar distances as controls (males: t 11 = 0.91; p = 0.38; females: t 11 = 2.02; p = 0.07). Regarding exploratory behavior, no differences between male WT and IP3R2 KO mice were observed for the number of rearings ( t 11 = 0.63; p = 0.54). Female IP3R2 KO display a similar number of vertical rearings when compared to WT mice ( t 11 = 1.32; p = 0.21). Next, we analyzed the distance swam by each mouse during 60 s (Figure 2.4B). Results showed that IP3R2 KO mice traveled similar distances, as compared with control littermates, for both males ( t 14 = 1.44; p = 0.17) and females ( t 14 = 1.35; p = 0.20). These findings support the use of adult IP3R2 KO mice for general behavior assessment in adulthood. Wire suspension Grasping Postural reflex Negative geotaxis Cliff aversion Labyrinthine reflex, coordination and strength Surface righting Air righting
66 Figure 2.4 – Adult IP3R2 KO mice display normal locomotor and exploratory functions. (A) Distance traveled and number of rearings recorded in the Open Field test. (B) Distance freely swam in a swimming pool during 60 s. WT mice are represented in black and IP3R2 KO mice are represented in red (males, left column; females, right column). Data plotted as mean ± SEM. n=6-8 per group. Discussion This comprehensive evaluation of somatic parameters and neurological reflexes shows that the constitutive deletion of IP3R2 does not interfere with the normal development of these mice. This is an important observation since this approach is being used by multiple laboratories to study the role of astrocytes in health and disease contexts, resulting occasionally in negative or controversial results (Oliveira et al. 2015; Guerra-Gomes et al. 2017). The IP3R2 KO model described in 2005 (Li et al. 2005) and used by most of the laboratories is a constitutive knockout mouse. Such gene deletions may result in developmental impairments and compensatory mechanisms that could bias experimental results obtained at different ages: development and adulthood. Until now, no study addressed the influence of IP3R2 genetic deletion during the postnatal period. Therefore, we performed a detailed evaluation of neurodevelopmental milestones that consisted of a thorough examination of the somatic parameters and neurological reflexes. Our results showed that IP3R2 KO mice displayed a body weight increase similar to controls, both in males and females. This result is in line with previous studies showing that adult WT and IP3R2 KO mice display a similar body weight (Li et al. 2015; Pinto-Duarte et al. 2019). More so, anogenital distance is not altered in both IP3R2 KO males and females, indicating a normal sexual differentiation in this mouse model. This result is in accordance with our normal colony breeding observations and with a previous report indicating that this mouse strain displays normal reproductive behavior (Petravicz et al. 2008). Regarding eye opening, our results are in line with the typical age described for the acquisition of this developmental Open Field A Water Maze B
67 milestone, which is on average at PND13 (Heyser 2004). These evidences indicate that IP3R2 KO mice display a healthy physical maturation. A neurodevelopmental assessment is fundamental to determine and discard possible alterations that might influence brain function and behavior performance (Sousa et al. 2006). We examined the acquisition of several neurological parameters related to sensory-motor reflexes, locomotion, coordination and strength in both WT and IP3R2 KO mice (males and females). In our study, we found that WT and IP3R2 KO mice similarly acquired a mature response for sensory-motor and auditory reflexes. These responses are in accordance with the average age previously described for C57BL/6 mice (Heyser 2004; Kleiber et al. 2011). From these evidences, we expect IP3R2 KO mice to be able to perform behavioral tests that depend on the tactile, auditory and motor abilities. Another battery of developmental tests covered a progressive maturation of sensory and motor skills depending on the vestibular system maturation (Khan et al. 2004). In these tests, such as surface righting, air righting, postural reflex, negative geotaxis, and cliff aversion, no significant differences were detected between WT and IP3R2 KO mice from both genders. This reflects a normal acquisition of sensorimotor-related parameters by IP3R2 KO mice, which discards the existence of possible compensatory mechanisms occurring in vestibular system related functions during postnatal development. Accordingly, in the wire suspension and grasping tests, we found that WT and IP3R2 KO mice equally performed this test within the typical range for the appearance of these developmental milestones (Heyser 2004; Kleiber et al. 2011). These tests are related to the acquisition of balance control mechanisms, which are intimately dependent on visual afferent information (Mesquita et al. 2007). It is noteworthy that by backcrossing the original mouse (Li et al. 2005) to C57BL/6J, we eliminated the retinal degeneration and associated loss of visual capacity described to mice that carry Black Swiss background. Altogether, these results discarded any alterations that could be occurring during development in IP3R2 KO mice that could compromise results obtained from these mice. These data provide strong evidence for a normal somatic and neurological development of IP3R2 KO mice, and they represent the first attempt to assess the influence of the constitutive genetic deletion of IP3R2. Finally, we assessed the general motor and exploratory phenotype of adult WT and IP3R2 KO mice (male and female). Our approach consisted of two behavioral tests that allowed us to obtain several readouts directly related with locomotor and exploratory function measures. In the open field test, we analyzed two motor-related measures: distance traveled and rearings. The first parameter corresponds to a pure measure of motor ability, while the second one is more related with exploratory drive (Seibenhener and Wooten 2015). Next, we evaluated swimming ability of WT and IP3R2 KO, to explore further their locomotor
68 capabilities. Our results show that WT and IP3R2 KO mice of both genders behaved similarly in both tests. Taken together, these findings show that IP3R2 KO mice display a normal development, which results in regular motor and exploratory behavior in adulthood. Deletion of IP3R2 in mice emerged as a promising genetic tool to study astrocytic Ca2+ signaling (Petravicz et al. 2008; Takata et al. 2011; Navarrete et al. 2012). Advances in the analysis of astrocytic Ca2+ elevations showed that astrocytes display rather complex spatial and temporal properties, whose functional consequences are still far from fully understood (Volterra et al. 2014; Shigetomi et al. 2016). Therefore, the field clearly needs tools to tackle differentially global or focal Ca2+ elevations. Functionally, the IP3R2 KO mouse model is an excellent tool to study the roles of global Ca2+ elevations specifically in astrocytes as shown by different labs (Di Castro et al. 2011; Navarrete et al. 2012; Yu et al. 2018; GuerraGomes et al. 2018; Mederos et al. 2019). Recently, by employing a viral tool to genetically drive the continuous extrusion of cytosolic Ca2+ in astrocytes, Yu et al. (2018) yielded similar levels of impairment of global events, which confirms the functional relevance of the IP3R2 KO mice. This observation together with our present findings fully supports the use of the IP3R2 KO mouse for the study of astrocyte involvement in all developmental stages and adulthood. Author Contributions SGG designed and performed the experiments, analyzed the data and wrote the manuscript; ELC participated in experimental conception and design; VMS, DSMN, JFV participated in experimental conception and data analysis; NS, LP and JFO supervised the study, secured funding, analyzed the data and wrote the manuscript.
69 CHAPTER 3 Sónia Guerra-Gomes, Vanessa Morais Sardinha, Gabriela Tavares, Inês Caetano, Diana Sofia Marques Nascimento, João Filipe Viana, Martin Irmler, Mareike Bamberger, Johannes Beckers, Michal Korostynski, Andreia Teixeira-Castro, Nuno Sousa, Luísa Pinto, João Filipe Oliveira The role of astrocytic calcium signaling in cognitive function (Manuscript to be submitted) (2019)
70 Title The role of astrocytic calcium signaling in the cognitive function Authors Sónia Guerra-Gomes1,2, Vanessa Morais Sardinha1,2, Gabriela Tavares1,2, Inês Caetano1,2, Diana Sofia Marques Nascimento1,2, João Filipe Viana1,2, Martin Irmler4, Mareike Bamberger4, Johannes Beckers4, Michal Korostynski5, Andreia Teixeira-Castro1,2, Nuno Sousa1,2, Luísa Pinto1,2, João Filipe Oliveira1,2,3 Affiliations 1 - Life and Health Sciences Research Institute (ICVS), School of Medicine, University of Minho, 4710057 Braga, Portugal 2 - ICVS/3B’s - PT Government Associate Laboratory, Braga/Guimarães, Portugal 3 - IPCA-EST-2Ai, Polytechnic Institute of Cávado and Ave, Applied Artificial Intelligence Laboratory, Campus of IPCA, Barcelos, Portugal 4 - Institute of Experimental Genetics, Helmholtz Zentrum München, German Research Center for Environmental Health, Neuherberg, Germany; 5 - Department of Molecular Neuropharmacology, Institute of Pharmacology, Polish Academy of Sciences, Kraków, Poland; Corresponding author: João Filipe Pedreira de Oliveira (PhD), Life and Health Sciences Research Institute (ICVS), School of Medicine, University of Minho, Campus de Gualtar, 4710-057 Braga, Portugal, Tel: +351-253-604871, Fax: +351-253-604809, [email protected] ACKNOWLEDGMENTS The authors are grateful to Prof. Alfonso Araque and Prof. Ju Chen for sharing the mice line. They acknowledge J. D. Silva for his helpful comments on the statistical analysis. The authors also acknowledge funding from national funds through Foundation for Science and Technology (FCT) fellowships to SGG, VMS, IC, CEEC to DSMN and IF grants to LP (IF/01079/2014) and JFO (IF/00328/2015); Bial Foundation Grants 207/14 and 037/18 to JO, and 427/14 to LP; Northern Portugal Regional Operational Programme (NORTE 2020), under the Portugal 2020 Partnership Agreement, through the European Regional Development Fund (FEDER) (NORTE-01-0145-FEDER-000013); FEDER Funds, through the
71 Competitiveness Factors Operational Programme (COMPETE), and The National Fund, through the FCT (POCI-01-0145-FEDER-007038). Abstract Astrocytes modulate neuronal activity and synaptic transmission, thus contributing to learning and memory processing. These cells sense and respond to neuronal activity by elevating their intracellular calcium (Ca2+) levels. Intracellular Ca2+ derives from different sources and displays complex spatiotemporal properties. Here we assessed the role of inositol 1,4,5-triphosphate receptor type 2 (IP3R2)- dependent Ca2+ signaling in cognitive function by testing the performance of a mouse model lacking this receptor in hippocampal-dependent tasks. Our results show that IP3R2 KO mice display enhanced cognitive performance in tasks associated with spatial reference (MWM) and fear memory (CFC). Moreover, we found that interfering with global astrocytic Ca2+ signaling in astrocytes does not affect the morphology of CA1 pyramidal neurons, but promotes an increased percentage of immature spines at apical dendrites. Strikingly, we found an astrocytic structural atrophy that seems to be compensated by an increased density of GFAP+-cells in the same hippocampal sub-region. At the molecular level, we identified the transcription factor Foxo1 as the transcription factor controlling a higher number of differentially expressed genes in IP3R2 KO mice, namely the astrocyte-specific syndecan-2 ( Sdc2 ) and ezrin ( Ezr ). These genes are related with fine cytoskeleton modulation and dendritic spine formation and might influence synaptic plasticity and cognitive function. To test this hypothesis, we overexpressed FOXO1 specifically in hippocampal astrocytes (GFAP-mCherry-FOXO1) of naïve C57BL/6J mice. We observed that GFAP-mCherry-FOXO1 mice display enhanced fear memory, as previously observed for IP3R2 KO mice, pointing to a key role of astrocytic FOXO1 regulated-pathways in cognitive modulation. Altogether, this work provides evidence that interfering with global IP3R2-dependent Ca2+ mechanisms or overexpressing FOXO1 in astrocytes, favors Foxo1 transcriptional activity towards the expression of relevant genes to support cognitive performance. Keywords astrocyte, calcium signaling, IP3R2, hippocampus, cognitive function, Foxo1
78 analyzed if they fulfilled the following criteria: 1) presence of a defined soma within the CA1 pyramidal layer; 2) complete impregnation along the entire length of the dendritic arborization; 3) no morphological alterations due to deficient impregnation or truncated branches. For each selected neuron, apical dendritic branches were reconstructed using a motorized microscope controlled by the Neurolucida software (MBF Bioscience, USA) under 100x magnification. This analysis allowed the acquisition of several parameters such as total length, number of endings, nodes, Sholl analysis, and spine number and classification. Dendritic segments of 30 µm, in the proximal and distal portions of the apical dendrite were randomly selected to assess the percentage of each spine category (from more immature to mature: thin, mushroom, thick and ramified) in both WT and IP3R2 KO mice. Data extraction was performed using the NeuroExplorer software (MBF Bioscience, USA). Tissue preparation and 3D-reconstruction of astrocytes The astrocytic main structure was analyzed by performing 3D-reconstructions of GFAP-stained cells using the Simple Neurite Tracer plugin, from the Fiji software (http://fiji.sc/Fiji), as previously described by Tavares et al. (2017). Brains were immersed in 4% paraformaldehyde solution [PFA, 0.1M, pH 7.4, in phosphate saline buffer (PBS)] overnight and then transferred to a 30% sucrose/PBS solution at 4ºC. Brains were frozen by immersion in isopentane (BDH Prolabo; cooled in liquid nitrogen) in Neg-50 frozen section medium (Thermo Scientific, USA) and stored at -20ºC until cryostat sectioning. Hippocampal sections (20 µm thick) containing the dHIP of WT and IP3R2 KO mice were incubated with the primary antibody rabbit anti-GFAP (1:200; Dako, Denmark) at 4ºC overnight. In the next day, sections were rinsed in PBS and incubated with the respective secondary antibody Alexa Fluor 594 goat anti-rabbit (1:1000; Molecular Probes, Invitrogen, USA). After several washes with PBS, brain sections were incubated with DAPI (1:1000; Invitrogen, USA), for cell nuclei labeling, for 10 min. Coverslips were mounted using ImmuMountTM (Thermo Scientific, USA) and analyzed in the confocal microscope (FV1000, Olympus, Japan). Each Z-stack image (1 µm z-step interval) was obtained with a resolution of 640 x 640 px under a 60x oil magnification, with a field size of 211.51 x 211.51 µm. The following criteria were taken into account to select astrocytes for reconstruction: i) GFAP-stained structure is located in the region of interest (hippocampal CA1 stratum radiatum ) and ii) present a clearly identifiable DAPI-stained nucleus with a iii) complete labeled GFAP-structure, without truncated processes. From this analysis we gathered relevant structural features as total process length, number of processes, process thickness, and Sholl analysis.
79 Assessment of astrocyte density We determined the number of GFAP+ cells in hippocampal CA1 stratum radiatum and stratum pyramidale layers brain sections from WT and IP3R2 KO mice. For that, we performed immunofluorescence staining for GFAP, as previously described. Images were acquired with the Olympus Fluoview FV1000 confocal microscope (Olympus, Hamburg, Germany) using a 20x objective (resolution: 1024x1024 px; field size: 635.28 x 635.28 µm). The number of GFAP+ cells was assessed using the Fiji plugin “Cell counter”. Molecular analysis To identify molecular signaling pathways and transcription factors altered in our mouse model of global astrocytic Ca2+ dysfunction, western blot and microarray analysis were performed in the total hippocampus of IP3R2 KO mice and WT littermates. Western blot Western blot analysis was performed in brain samples from WT and IP3R2 KO mice. Total hippocampus was lysed using a cold HEPES-buffered sucrose (0.32M sucrose, 4mM HEPES pH 7.4) buffer with 25x protease inhibitors, 1% Nonidet-P40 and 0.5% sodium dodecyl sulfate (SDS). Lysed samples were sonicated and centrifuged at 10.000 rpm during 25 min at 4ºC. The collected supernatant was quantified for the total amount of protein using the Bradford protein assay (Bio-Rad, USA). Total lysates were denatured in 2x Laemmli Buffer (Bio-Rad, USA) by heating the samples at 98ºC for 5 min. Next, 50 µg of protein were loaded onto an SDS-PAGE gel (10%) and transferred to a nitrocellulose membrane (Transblot Turbo Kit, Bio-Rad, USA). Membrane was blocked for 1 h in a 5% non-fat milk/TBS solution, to prevent non-specific background binding of the primary and/or secondary antibodies. Next, the membrane was incubated with the primary antibody rabbit anti-GFAP (1:50.000; Dako, Denmark) overnight. The next day, the membrane was washed with TBS-T, and incubated, at room temperature with agitation during 2 h, with the respective secondary antibody anti-rabbit HRP (1:15.000; BioRad USA). After this period, the membrane was washed again three times in TBS-T, followed by the detection of the chemiluminescent signal using the Clarity Western ECL substrate kit (Bio-Rad, USA) using a gel blotting imaging system (Chemidoc, Bio-Rad, USA). Protein expression quantification was performed using the Image Lab software (Bio-Rad, USA.) All the samples were normalized for a loading control protein (mouse anti-α-tubulin; 1:500, DSHB, USA).
80 Transcriptome analysis Microarray analysis was performed in total RNA extracted from the whole hippocampus of WT and IP3R2 KO mice (n=3 biological replicates/group) that did not perform any behavioral test. A single experienced researcher performed hippocampus macrodissection, in order to avoid variability. After dissection, tissues were immediately frozen in dry ice and stored at -80ºC until RNA extraction. RNA isolation Total RNA was extracted using QIAzol (Qiagen, Germantown, MD, USA) in combination with the miRNeasy Mini kit (Qiagen, Germantown, MD, USA). The Agilent 2100 Bioanalyzer was used to assess RNA quality and only high quality RNA (RIN > 7) was used for microarray analysis. Expression profiling Total RNA (40 ng) was amplified using the Ovation Pico WTA System V2 in combination with the Encore Biotin Module (NuGEN Technologies, Inc, San Carlos, CA, USA). Amplified cDNA was hybridized on a mouse Gene 2.0 ST arrays (Affymetrix, Santa Clara, CA, USA). Staining and scanning (Scanner 3000 7G) was done according to the Affymetrix expression protocol including minor modifications as suggested in the Encore Biotion protocol (NuGEN Technologies, Inc). cDNA synthesis and real-time quantitative PCR analysis Total RNA (500 ng) was converted from the samples used for microarray analysis (total hippocampus) using qScript cDNA SuperMix (Quanta Biosciences, Gaithersburg, MD, USA). The primers for the selected genes of interest for microarrays confirmation were designed using PRIMER-BLAST (NCBI, http://www.ncbi.nlm.nih.gov/tools/primer-blast/) (Table 3.1).
81 Table 3.1 - Forward and reverse sequences of oligonucleotide primers of the selected genes used for microarrays data validation. Gene Forward Sequence (5’→3’) Reverse Sequence (5’→3’) F5 CCT GGC AAG CCA AGG CAA ACA GCC ACA CCC TGG TCA CTG TAC T Sulf1 TGA GTG CTT GAG GAC GTG TTT C TGG CCC TCA GCA CCT GAA AAT AC Clic6 CCT GGG ACG AAC CCT CCT TTC A AGG GTA CCT CGG GGG AAC TAA CT Vat1l GGC CCT GGA GGA GGT AAA AGA TCT TCC CAG CCA TCC TCC ACT TT Vamp8 GGA AGC CAC GTC TGA ACA CTT CAA AGA GGG CTC CTC TTG GCA CAT A Kl ACT ACG TTC AAG TGG ACA CTA CTC GGC CGG ATG GCA GAG AAA TCA Gper1 AAC GCC ACG GCA CAG ATC A TGG TGG GTG CAT GGC AGA AAT G Grik1 ACA TTG AGC AGT GTC TCT CTT TCA GGG TCA CGC CAC AGT CTC TT Plcb3 GCC GGG CCA GAT CCA GTA AA GCG AAT GGA CTC ACT TCG GTT GAA S100a4 TCT TGG TCT GGT CTC AAC GGT TA TAG GCA GCT CCC TGG TCA GTA Cib2 GAC AAC TAC CAG GAC TGC ACT TT CCT CGG AGA AAG CCT CCA CAA TC Smad3 GAG GAG AAG TGG TGC GAG AAG CAG TGA CCT GGG GAT GGT AAT G Quantifications were performed in a Fast Real-Time PCR System (Applied Biosystems, USA) using the 5x HOT FIREPol® EvaGreen® qPCR Mix Plus, ROX (Solis Biodyne, Estonia). The housekeeping gene 18S rRNA was used as an internal control and the relative expression was calculated using the ΔΔCt method. Bioinformatic analysis Functional annotation analysis We analyzed enriched molecular functions associated with IP3R2 KO mice differentially expressed targets using PANTHER Classification System (http://www.pantherdb.org/). After importing our list of genes to the PANTHER tool, we selected a functional classification view in pie charts, and selected ontology “Molecular function”. Prediction of transcription regulators To identify possible regulons from our list of differentially expressed genes, the Cytoscape plugin (https://cytoscape.org/) – iRegulon – was used (Janky et al. 2014). This plugin detects transcription factors (TFs) and its direct regulated targets based on motif discovery from a list of genes. Briefly, we imported the network file containing our gene list into Cytoscape. After selecting all nodes and edges, we
82 performed an iRegulon analysis using the following ranking options: species and gene nomenclature (Mus musculus, MGI symbols), motif collection [10K (9713 PWMs)], no track collection, putative regulatory region (20kb centered around TSS), motif rankings database [20 kb centered around TSS (7 species)], track rankings database [20 kb centered around TSS (ChIP-seq-derived)]. Regarding recovery, the following parameters were applied: enrichment score threshold ≥ 3.0 and maximum false discovery rate (FDR) on motif similarity = 0.001. Prediction of functional interactions Target genes regulated by the most enriched TF from our iRegulon analysis – Foxo1 – were uploaded at the STRING tool (https://string-db.org/, version 11.0). These genes encode proteins that are involved in several cellular functions. STRING analysis retrieved possible functional associations between our targets. The minimum required interaction score established was 0.4 (medium confidence). We combined these results with a published transcriptome database (https://web.stanford.edu/group/barres_lab/brain_rnaseq.html; http://www.brainrnaseq.org/) (Zhang et al. 2014), in order to identify Foxo1 targets in IP3R2 KO mice that have functional similarity and that are astrocyte-enriched genes. Viral approach to overexpress FOXO1 in hippocampal astrocytes Stereotaxic virus injection AAV5-GFAP(0.7)-mCherry-2A-m-FOXO1 was obtained from Vector Biolabs (USA) and its titer corresponds to 3.9 x 1012 genome copies (GC)/mL. The GFAP (gfaABC1D) promoter drives the expression of both FOXO1 and mCherry, which have a 2A linker in between for protein co-expression. Male C57BL/6J mice, 8-12 weeks old were anesthetized with a mixture of ketamine (75 mg/kg, i.p.; Imalgene 1000, Merial, United States) and medetomidine (1 mg/kg, i.p.; Dorbene Vet, Pfizer, United States), injected locally in the head with lidocaine (30 µL/0.5% lidocaine) before incision and submitted to a stereotaxic surgery for the bilateral injection of either saline (Sham group, n=8) or an adenoassociated virus 5 (AAV5) under the control of a GFAP promotor to specifically overexpress FOXO1 in astrocytes (n=13). AAV5-GFAP(0.7)-mCherry-2A-m-FOXO1 (1 µL; 3.9 x 1012 GC/mL) was bilaterally injected into the dHIP (coordinates from bregma, according to Paxinos and Franklin (2004): 1.8 mm anteroposterior, 1.3 mm mediolateral, and 1.3 mm dorsoventral). The injection volume and flow rate were established at 100 nL/min. At the end of the procedure, the needle was pulled up 0.1 mm and it was left in place during approximately 5 min to allow proper viral vector diffusion. After this, mice were
83 removed from the stereotaxic apparatus and sutured. To revert the effect of anesthesia, 1 mg/kg of atipamezole was administered. After surgery, mice received subcutaneous injections of an antiinflammatory (Carprofeno, 5 mg/kg), an opioid for analgesia (Bupac®, 0.05 mg/kg) and vitamins (Duphalyte/Pfizer, USA). Animals were allowed to recover during 4 weeks prior to behavioral testing. Immunofluorescence for the confirmation of virus expression Two and 12 weeks post injection, AAV5-GFAP(0.7)-mCherry-2A-m-FOXO1 mice (2 weeks, n=1; 12 weeks, n=2) were anesthetized with a mixture of ketamine and medetomidine and transcardially perfused with 0.9% saline, followed by perfusion with 4% PFA in PBS. Brain was removed and kept in 4% PFA overnight at 4ºC. In the next day, it was washed in PBS 1x and cryoprotected in 30% sucrose/PBS 1x solution. Free-floating sections (50 µm thick) were obtained using a vibrating-blade microtome Leica VT 1000 S. Sections were washed in PBS 1x and PBS-Tx 0.3% and incubated during 1 h with a blocking solution containing PBS 1x/10% FBS/0.25 M glycine. These steps were followed by the incubation with primary antibodies overnight at 4ºC: chicken anti-mCherry (1:1000, HenBiotech, Portugal), rabbit anti-NeuN (1:100, Cell Signaling, USA), rabbit anti-FOXO1 (1:100, Cell Signaling, USA), mouse anti-GFAP (1:200, Sigma Aldrich, USA), and rabbit anti-S100β (1:200, DakoCytomation, USA). In the next day, sections were washed in PBS 1x and incubated during 2 h at room temperature with the secondary antibodies: Alexa Fluor 594 goat anti-chicken, Alexa Fluor 488 goat anti-rabbit, Alexa Fluor 647 donkey anti-mouse (1:1000, Thermo Fisher Scientific). Finally, sections were washed again in PBS 1x, incubated during 10 min with DAPI (1:1000), rinsed in PBS 1x twice and were mounted in Immu-MountTM (Thermo Scientific, USA). Images were acquired in an Olympus Fluoview FV100 confocal microscope (Olympus, Germany) using the 20x objective and 100x oil immersion objective. Statistical analysis Statistical significance was considered for p < 0.05. All data assumed a Gaussian distribution as revealed by the central limit theorem for samples n > 30 (MWM data) or by the Kolmogorov-Smirnov normality test. Two-tailed Student’s t-test was applied for comparisons between WT and IP3R2 KO or Sham and GFAP-mCherry-FOXO1 mice, whereas a Two-way analysis of variance (Two-way ANOVA), followed by Sidak post hoc analysis was used for multiple comparisons. For the analysis of different behavioral categories of strategies in the MWM, the two-sided Chi-square test was carried out. All statistical analyses were carried out using SPSS 22.0 or GraphPad Prism 7.04.
84 For transcriptome analysis, Pearson coefficients were calculated to assess the correlation between microarray and RT-PCR data. Expression console (v.1.4.1.46, Affymetrix) was used for quality control and to obtain annotated normalized SST-RMA gene-level data (standard settings including median polish and sketch-quantile normalization). Statistical analyses were performed with the statistical programming environment R [R Development Core Team (2005)]. Genewise testing for differential expression was done employing the limma t-test and Benjamini-Hochberg multiple testing correction (FDR < 10%). Heatmaps were generated with Genesis (Sturn et al. 2002) and cluster dendrograms with the R script hclust. Results A previous neurodevelopmental evaluation of IP3R2 KO mice showed that male and female IP3R2 KO mice display normal development. Therefore, the IP3R2 KO mouse model is reliable to study the functional impact of global IP3R2-dependent astrocytic Ca2+ elevations in adulthood (Chapter 2). We performed a full behavioral characterization in adult WT and IP3R2 KO mice by assessing several behavioral dimensions: anxiety, mood, and cognition (Figure S 3.1A-F). Anxious-like behavior was assessed by the Elevated-Plus Maze (EPM), Open Field (OF) and Light-Dark box (LD box). Impulsivity was evaluated by the Acoustic Startle Test. Depressive-like behavior was assessed by the Forced Swim Test (FST) and Tail Suspension Test (TST). Our results show that IP3R2 KO mice do not present an anxiousor depressive-like phenotype that could bias cognitive performance. Thus, we evaluated the impact of IP3R2dependent Ca2+ signaling in hippocampal-dependent cognitive tasks. IP3R2 KO mice display an enhanced performance in hippocampal-dependent cognitive tasks We employed the IP3R2 KO mouse model of global astrocytic Ca2+ signaling dysfunction to study its implications for cognitive function, mainly in hippocampal-dependent tasks. The MWM task was used to assess spatial reference memory. In this paradigm, mice had to learn the hidden platform location, which was kept always in the same position, based on their spatial navigation abilities. Results show that both WT and IP3R2 KO mice learned to find the platform and improved their performance during the four days of testing similarly (Figure 3.1A; escape latency: day, F 3,207 = 128.70; p < 0.0001; distance swam: day, F 3,207 = 74.58; p < 0.0001), but IP3R2 KO mice performed better in day 3 ( t 276 = 3.00; p < 0.05). Moreover, in the probe trial, both genotypes displayed a similar preference for the quadrant where the platform was hidden (Figure 3.1B). Interestingly, the analysis of the strategies used to reach the platform during the acquisition phase revealed that IP3R2 KO mice employed more hippocampal-dependent strategies as compared with WT mice, indicating an enhanced spatial orientation (Figure 3.1C). In fact, WT and IP3R2
85 KO mice display similar percentages of failure to reach the platform, but IP3R2 KO mice use less random and more directed strategies when compared to control mice (Figure 3.1C; χ2 (1) = 9.59; p < 0.01). Figure 3.1 – IP3R2-dependent Ca2+ signaling modulates spatial reference memory in the MWM. (A-C) Spatial reference memory task (RM) of the Morris Water Maze (n=32-39 per group). (A) RM task scheme (top) with the respective learning curves (bottom) depicting escape latencies and distance swam to reach the platform during the 4 days of testing, for WT and IP3R2 KO mice. (B) Probe trial (top, right) showing the percentage of time spent in the goal quadrant for both genotypes. (A, B) WT mice are represented in black and IP3R2 KO mice in red bars. Data plotted as mean ± SEM. (C) Representative scheme of the swimming strategies (top) used to reach the platform. Swimming tracks were classified as failure (gray), random scanning (orange) or directed to the platform (green). Search strategy used by WT and IP3R2 KO mice (bottom, left) over the 16 trials of RM. Graphical representation showing the proportion of failures and strategies used to reach the platform (bottom, right) for WT and IP3R2 KO mice. ** orange, refers to difference in random strategies used between WT and IP3R2 KO mice, ** green, refers to difference in directed strategies used between WT and IP3R2 KO mice. * p < 0.05 ** p < 0.01.
86 To explore further the role of global astrocytic Ca2+ signaling in a different form of hippocampal-dependent memory, we tested WT and IP3R2 KO mice in the CFC task. In the training day, we found similar baseline activity and freezing responses, after the application of three footshocks associated with a light cue (Figure 3.2A). These results discarded any genotype-related alterations in baseline activity and similar conditioning for WT and IP3R2 KO mice. Moreover, we observed that the light-shock pairings triggered a conditioned fear response in both genotypes (session, F 1,35 = 433.2; p < 0.0001). In the next day, reexposure to the same context revealed a significant increase in freezing response by IP3R2 KO mice, as compared with WT littermates (Figure 3.2B; t 35 = 2.22; p = 0.03). In the cue probe phase (day 3), no differences between genotypes were observed after light stimulus presentation (Figure 3.2C). Overall, these results support an impact of IP3R2-dependent Ca2+ signaling in cognitive functions, specifically enhancing the performance in hippocampal-dependent tasks. These results prompted us to look for histological and molecular correlates to better understand the underpinnings of this cognitive enhancement. Figure 3.2 – IP3R2 KO mice display an increased context-dependent fear memory. (A-C) WT and IP3R2 KO mice were tested in a contextual fear conditioning paradigm (n=16-21 per group). (A) Training session depicting baseline activity (left) and freezing acquired after light-shock exposure (right) at context A. (B) Re-exposure to context A (left) enhanced freezing response in IP3R2 KO mice, compared to WT animals, but revealed similar responses after switching to a new context (context B, right). (C) Presentation of a light stimulus in context B led to similar freezing responses between genotypes. WT mice are represented in black and IP3R2 KO mice in red bars. Data are presented as mean ± SEM. * p < 0.05, **** p < 0.0001.
87 IP3R2 KO mice display intact dendritic morphology, but an increased percentage of immature spines in dorsal CA1 pyramidal neurons To correlate the behavioral outcomes observed in IP3R2 KO mice with morphological alterations, we performed a 3D-reconstruction of dorsal CA1 hippocampal pyramidal neurons (Figure 3.3A). These neurons are responsible for the integration of spatial, contextual and emotional information, and are actively involved in the processing of cognitive function (Graves et al. 2012). Our results showed that WT CA1 neurons are morphologicaly similar to IP3R2 KO neurons for the parameters assessed (Figure 3.3B). Specifically, no differences were found in total dendritic length, number of endings and nodes. Moreover, Sholl analysis revealed unaltered dendritic arbor complexity of IP3R2 KO neurons as compared with WT neurons (Figure 3.3C). Since dendritic spine structure is crucial for long-term memory, we further analyzed the relative distribution of different spine subtypes to understand their dependence on astrocytic Ca2+ signaling (Kasai et al. 2010). For that, we performed a spine categorization (thin, mushroom, thick Figure 3.3 – CA1 pyramidal neurons in the dHIP of IP3R2 KO mice are morphologically similar to those of WT, but display more immature spines. (A) Representative 3D reconstructions of dHIP CA1 pyramidal neurons from WT and IP3R2 KO mice (scale bar = 100 µm). (B) Morphological parameters of apical dendrites, namely total length, number of endings and nodes (from left to right) (n=22-27 neurons per group). (C) Sholl analysis for intersections of apical dendrites. (D) Percentage of each spine type at apical proximal and distal dendrites (n=17-19 neurons per group). WT mice are represented in black dots and lines/bars and IP3R2 KO mice in red dots and lines/bars. Data plotted as mean ± SEM. **** p < 0.0001.
94 weeks; Figure 3.7B). At a higher magnification, we confirmed the well-distributed expression throughout CA1 stratum oriens down to the stratum radiatum layer (Figure 3.7C). Specific antibodies were used to double-label mCherry in combination with astrocytic (GFAP and S100β) or neuronal (NeuN) markers (Figure 3.7D-F) in brain tissue collected only 2 weeks after the viral injection. Our results show that mCherry co-localizes with both S100β + (Figure 3.7D) and GFAP+ (Figure 3.7E) cells, which confirms the specificity of the promotor GFAP(0.7) for astrocytes. Additionally, we excluded any neuronal expression of the virus, as shown by the immunostaining for the neuronal-specific nuclear protein marker, NeuN, which does not overlap with mCherry staining (Figure 3.7F).
95 Figure 3.7 – AAV5-GFAP(0.7)-mCherry-2A-m-FOXO1 effectively infected astrocytes of the dHIP CA1 of C57BL/6 mice. (A) Representative scheme depicting the bilateral stereotaxic injections of AAV5-GFAP(0.7)-mCherry-2A-m-FOXO1 virus in the hippocampal CA1 area of C57BL/6 mice. (B) Anti-mCherry reporter fluorescence (red) and DAPI staining in the hippocampus. (C) 12 weeks post-injection it is possible to observe a clear infection of CA1 stratum oriens (or) and a spreading of the virus through the stratum pyramidale (pyr) and stratum radiatum (rad) layers. (D-F) Confocal micrographs of immuno-histochemistry in brain slices illustrating co-expression of mCherry reporter with astrocytic (S100β and GFAP), but not neuronal (NeuN) markers. (D,E) mCherry (red) reporter gene is coexpressed with S100β (D, green) and GFAP (E, green). (F) mCherry (red) expression is absent in NeuN+ neurons (green). White arrows indicate co-localization of (D) mCherry/S100β and (E) mCherry/GFAP and the (F) absence of a co-localization of mCherry with NeuN. DAPI staining, blue. Scale bars depicted in representative images.
96 Furthermore, we evaluated the co-localization of FOXO1 expression with mCherry (Figure 3.8A) and GFAP (Figure 3.8B) immunostainings. We confirmed that FOXO1 is overexpressed specifically in astrocytes already at two weeks after infection with AAV5-GFAP(0.7)-mCherry-2A-m-FOXO1. Figure 3.8 – Astrocyte-specific overexpression of FOXO1 in the dHIP CA1 of C57BL/6 mice with AAV5-GFAP(0.7)-mCherry-2A-m-FOXO1. (A-B) Confocal images of immuno-histochemistry in brain slices showing that FOXO1 (green) is co-expressed with mCherry reporter (red) and the astrocytic marker GFAP (white). (A) FOXO1 (green) expression is detected in mCherry+-cells (red) and (B) co-localizes specifically with the astrocyte marker GFAP. White arrows indicate colocalization of (A) mCherry/FOXO1 and (B) mCherry/FOXO1/GFAP. DAPI staining, blue. Scale bars depicted in representative images.
97 Based on our previous findings, Foxo1 regulates the expression of key astrocytic genes that appear to support hippocampal-dependent cognitive function, in IP3R2 KO mice. Therefore, we tested whether increasing FOXO1 expression in the hippocampus of pure C57BL/6 mice would recapitulate the memory enhancement observed in IP3R2 KO mice. For that, we started by performing a behavioral characterization of Sham and GFAP-mCherry-FOXO1 mice four weeks after viral injection. We assessed anxious-like behavior, motor, and exploratory functions. We observed that overexpression of FOXO1 specifically in astrocytes induced neither an anxious-like phenotype nor alterations in locomotor activity (Figure S 3.2). Next, we evaluated contextual fear memory in both groups, using exactly the same paradigm and conditions previously described for WT and IP3R2 KO mice (Figure 3.9). On the first day of testing, both groups displayed a similar freezing response before and after light-shock pairings (Figure 3.9A). From this analysis, we excluded any effect of the injection and concluded that both Sham and GFAP-mCherry-FOXO1 mice had similar freezing responses after the conditioning trials. Furthermore, light-shock pairings triggered a conditioned fear response in both groups (session, F 1,17 = 352.2; p < 0.0001). Interestingly, re-exposure to the same context 24 h later revealed an increased freezing response of GFAP-mCherryFOXO1 mice as compared with Sham (Figure 3.9B; t 17 = 2.918, p = 0.0096). This result goes in line with our previous observations using IP3R2 KO mice regarding contextual memory. Moreover, switching both groups to a new environment 2 h later, decreased freezing response percentages in both groups, but GFAP-mCherry-FOXO1 mice still presented an increased freezing response, as compared with Sham animals (Figure 3.9B; t 17 = 3.221, p = 0.005). Nevertheless, in the cue probe, when the light stimulus was presented in a new context, similar freezing responses were observed for both groups. Altogether, these results support the role of astrocytic FOXO1 in the control of cognitive function, namely by enhancing hippocampal contextual memory.
98 Figure 3.9 – GFAP-mCherry-FOXO1 mice present an increased contextual fear memory. (A-C) Sham and GFAP-mCherry-FOXO1 mice were tested in a contextual fear conditioning paradigm (n= 7-12 per group). (A) Training session depicting baseline activity (left) and freezing acquired after light-shock exposure (right) at context A. (B) Re-exposure (24 h later) to context A (left) enhanced freezing response in GFAP-mCherry-FOXO1 mice. After switching to a new context, GFAP-mCherry-FOX01 mice still displayed a higher freezing response, as compared with Sham mice (context B, right). (C) Presentation of a light stimulus in the new context led to similar freezing responses between groups. Sham mice are represented in black and GFAP-mCherry-FOXO1 mice in red bars. Data are presented as mean ± SEM. ** p < 0.01, **** p < 0.0001. Discussion In this work, we observed that the deletion of IP3R2-dependent astrocytic Ca2+ signaling regulates cognitive performance. Particularly, we demonstrated that this regulation is mostly observed in hippocampaldependent tasks. In this region, CA1 pyramidal neurons that are crucial for this performance, display an intact dendritic tree, but have more immature spines. We also showed that IP3R2 KO astrocytes display a shrinkage of its main structure, which appears to be compensated by an increased number of GFAP+ cells. At a molecular level, we identified the transcription factor Foxo1 as a key regulator of 76 differentially expressed genes in the IP3R2 KO mouse model. Looking deeper into the functional interactions between Foxo1 target genes and its cell-type enrichment from a widely accepted database, we found two genes, Sdc2 and Ezr, which are mainly transcribed in astrocytes. They are described to regulate spine structure, with impact to synaptic function and therefore expected to play an important role in behavior dependent on the hippocampus of IP3R2 KO mice. We will discuss these results in light of the existing literature, following the hypothesis depicted in Figure 3.10.
99 Figure 3.10 – Schematic representation of the hypothesis linking our molecular, structural and behavioral observations. We studied the IP3R2 KO mouse model, since it is a widely used model of astrocyte signaling dysfunction described in the literature. Besides, IP3R2 KO mice are reported to maintain intact Ca2+ levels in CA1 pyramidal neurons, whereas intracellular Ca2+ elevations in the soma and main astrocytic processes are abolished (Petravicz et al. 2008; Navarrete et al. 2012). Astrocytes from this model still present “focal” astrocytic Ca2+ elevations in far distant microdomains (Srinivasan et al. 2015), mainly arising from the mitochondria (Agarwal et al. 2017), which are expected to play a role in the control of specific synapses
100 or micro-circuits. Recently, the Bal Khakh’s laboratory used a viral approach to express Ca2+ pumps under the control of a GfaABC1D promoter to continuously extrude this ion in striatal astrocytes (Yu et al. 2018). In this study, authors found that this approach mimics the decreased frequency of spontaneous Ca2+ signals and reduced Ca2+ signal amplitude in astrocyte branches, as observed in IP3R2 KO mice. Both models responded similarly to phenylephrine, although IP3R2 KO mice displayed a larger reduction of Ca2+ responses. Together with the available Ca2+ imaging observations (Navarrete et al. 2012), these data confirm that IP3R2 KO astrocytes are not able to use reticular Ca2+ to drive Ca2+-dependent pathways. The forkhead box O (FOXO) family of transcription factors is responsible for the regulation of several genes involved in distinct cellular processes, with possible roles in the modulation of cognitive function (Eijkelenboom and Burgering 2013; Kim and Webb 2017). Four members compose the FOXO family: FOXO1, FOXO3, FOXO4, and FOXO6. It is accepted that FOXOs are regulated via two main conserved signaling pathways: the insulin/insulin-like growth factor/protein kinase B (AKT) signaling, in response to growth factors; and the Jun N-terminal kinase signaling, upon stress conditions. These pathways regulate FOXO transcriptional activity through post-translational modifications (PTM), namely phosphorylation. However, it is known that a higher complexity of signaling pathways and PTMs regulate FOXOs activity (Eijkelenboom and Burgering 2013). For instance, FOXOs can be ubiquitylated, methylated and acetylated, affecting both its subcellular location and its activation state (Calnan and Brunet 2008; Eijkelenboom and Burgering 2013). However, to the best of our knowledge the mechanisms underlying FOXOs regulation in astrocytes are unknown. According to a widely used transcriptomic database (Zhang et al. 2014), Foxo1 is highly expressed in astrocytes, as compared with the remaining family members. Interestingly, the available literature reports that in different cell types (Kawakami et al. 2004; Barragan et al. 2006; Wu et al. 2009) protein kinase C (PKC) controls AKT activation, which in turn is responsible for the negative regulation of Foxo1 activity (Biggs et al. 1999; Kim and Webb 2017). Since PKC is tightly regulated both by DAG and cytosolic Ca2+, in the absence of Ca2+, Foxo1 activity is disinhibited. Our results showed that in the hippocampus of IP3R2 KO mice, Foxo1 induces the expression of a large number of targets. Among these targets, we found astrocyte-enriched genes such as Sdc2 and Ezr (Zhang et al. 2014), which play a role in perisynaptic astrocyte processes (PAPs) morphology and dendritic spine formation (Lin et al. 2007; Lavialle et al. 2011; Molotkov et al. 2013; Hu et al. 2016) and might contribute to the observed shift to more immature spines, leading to changes in synaptic function with impact to cognitive behavior. Besides the spine modulation, Ezr is found at PAPs and is involved in morphological plasticity during synaptic communication (Derouiche and Frotscher 2001; Lavialle et al. 2011; Heller and Rusakov 2015). It is described that PAPs correspond to areas of proximity between dendritic spines and
101 astrocytic processes and display structural plasticity that is fundamental for learning and memory. Moreover, activity-related structural remodeling of PAPs was shown to be dependent on astrocytic Ca2+- elevations (Perez-Alvarez et al. 2014). One should not rule out that, at least partially, by displaying IP3R2independent Ca2+ events in the microdomains, IP3R2 KO astrocytes might circumvent the lack of global Ca2+ signals in the soma and main processes. The CA1 stratum radiatum of IP3R2 KO present a higher density of astrocytes with simpler GFAP+ arbor complexity. By evaluating the GFAP-stained structure of astrocytes, we targeted its main structure, which corresponds to around 15% of astrocytes total volume (Bernardinelli et al. 2014). The astrocytic atrophy observed seems to be a result of global Ca2+ signaling interference. In fact, a study using the IP3 “sponge” mouse model reported a reduction in the astrocytic coverage in hippocampal CA1 stratum radiatum (Tanaka et al. 2013). However, our observations show that an increased cellular density of GFAP+ in the stratum radiatum layer circumvents astrocytic atrophy. This screening of astrocyte structure lacks however, the detail required to disentangle the fine astrocyte morphology, which could provide additional cues on further morphological consequences. To assess the detailed astrocytic morphology, including its microdomains, it would require much more powerful and time-consuming monitoring methods (Heller and Rusakov 2015). Altogether, these data suggest that the individual or combinatory effects of (1) a higher density of less complex astrocytes at the CA1 stratum radiatum , (2) Sdc2 and Ezr modulation of PAPs and vicinal spines shall have an influence on synaptic plasticity. Synaptic plasticity is frequently observed in mouse models of cognitive enhancement (Lee and Silva 2009). Our data shows that interfering with IP3R2-dependent Ca2+ signaling contributes to a shift to more immature (thin) spines in apical dendrites of CA1 pyramidal neurons, critical for cognitive processing. Thin spines are recognized as “learning spines”, whereas mushroom spines are called “memory spines” (Bourne and Harris 2007) mainly due to their intrinsic features. Thin spines display a more plastic and transient structure and retain biochemical signals such as Ca2+, which favors learning mechanisms (Bourne and Harris 2007). Moreover, glia interacts closer to thin spines associated with the learning process (Medvedev et al. 2014). A dendritic spine simplification, from large “memory” spines to thinner “learning” spines, is associated with LTD induction (Bourne and Harris 2007). Evidences point to an important role of hippocampal LTD in memory formation and consolidation (Nicholls et al. 2008; Ge et al. 2010; Dong et al. 2013b). Chen et al. (2013) described that astrocytes contribute to heterosynaptic long-term depression (hLTD) in the hippocampus. However, interfering with astrocytic Ca2+ signaling impairs this phenomenon. Recently, Navarrete et al. (2019) found that astrocytes control NMDARdependent LTD, supporting memory enhancement. Specifically, these authors show that the deletion of
102 p38α MAPK from hippocampal astrocytes impairs LTD, which has an enhancing effect in hippocampalassociated fear memory. Moreover, a recent work provided new insights into NMDAR-dependent LTD in IP3R2 KO mice under a C57BL/6 background (Pinto-Duarte et al. 2019). This study shows that IP3R2 KO mice present an impaired hippocampal LTD. This observation, together with the Navarrete et al. (2019), supports the hypothesis that LTD impairment in IP3R2 KO mice could underlie the cognitive enhancement observed in our experiments. The influence of global IP3R2-dependent Ca2+ signaling in behavior is still poorly understood. We performed a full behavioral characterization of adult IP3R2 KO mice on a C57BL/6J background. For a behavioral analysis, it is important to take into consideration the mouse strain and the targeted mechanism since it could easily mislead data interpretation (Oliveira et al. 2015; Guerra-Gomes et al. 2017). In accordance with other studies, we found that IP3R2 KO mice do not display an anxious phenotype (Cao et al. 2013; Tanaka et al. 2013; Petravicz et al. 2014; Pinto-Duarte et al. 2019). However, in what concerns depressive-like behavior results are contradictory. IP3R2 KO mice generated on a Black Swiss background display increased learned helplessness (Cao et al. 2013), in opposition to an IP3Rs knocking-down approach, which leads to an antidepressant phenotype (Galeotti et al. 2008). Moreover, the specific deletion of IP3R2 in astrocytes neither affects depressive-like behavior nor impulsivity (Petravicz et al. 2014), as we have observed, suggesting that the previously described anxiousor depressive-like phenotypes may be dependent on the mouse genetic background. Regarding the normal motor function and coordination, our observations are in agreement with other reports targeting IP3R2-dependent Ca2+ signaling (Cao et al. 2013; Tanaka et al. 2013; Petravicz et al. 2014; Kim et al. 2016; Pinto-Duarte et al. 2019). We observed a cognitive enhancement in the performance of hippocampal-dependent tasks by IP3R2 KO mice. Interestingly, an enhanced cognitive performance was previously described in other mouse models targeting Ca2+-related mechanisms (Lee and Silva 2009). For instance, the ryanodine receptor 3 KO mice display an increased spatial learning performance (Futatsugi et al. 1999). Authors claim that this effect might be the result of smaller intracellular Ca2+ increases in neurons that are sufficient to induce LTP and modulate cognitive performance. In a different study, a mouse model knockout for S100B, a Ca2+ binding protein and highly expressed in astrocytes (Wang and Bordey 2008), presents a cognitive performance similar to what we have observed in the IP3R2 KO mice. S100B KO mice display an increased spatial reference memory in the MWM and increased fear memory (Nishiyama et al. 2002). Moreover, Jeon et al. (2003) reported that genetic deletion of plasma membrane Na+/Ca2+ exchanger 2 increased the time required for neuronal-induced intracellular Ca2+ clearance and resulted in enhanced performance on tasks
103 engaging the hippocampus. Ca2+ signals play a pivotal role in functional neuron-astrocyte interactions (Losi et al. 2017). Together, these observations indicate that cytosolic Ca2+, and its downstream pathways, are critical for cognitive function. Overall, our findings show that interfering with global Ca2+ signaling in astrocytes favors cognitive performance in hippocampal-dependent tasks. We found that Foxo1 is particularly active leading to the increased expression of astrocyte-specific Sdc2 and Ezr . These might contribute to alterations in fine cytoskeleton and dendritic spine formation in the hippocampus, with impact in synaptic plasticity, leading to a cognitive enhancement. To test this hypothesis, we used a viral approach to specifically overexpress FOXO1 in hippocampal astrocytes of naïve C57BL/6 mice. This specific modulation was enough to recapitulate the enhanced fear memory previously observed in IP3R2 KO mice, confirming an astrocyte role in cognitive modulation through FOXO1 regulated-pathways. Our results suggest that astrocytic Ca2+ regulates Foxo1 transcriptional activity, which ultimately controls cognitive function via morphological changes and synaptic plasticity. The identified gene candidates should be further tested as therapeutic targets for cognitive disabilities, for instance in aging, neurodegenerative disorders or stress-related pathologies. Author Contributions SGG designed the study, performed and analyzed the experiments and wrote the manuscript; VMS, GT, IC, JFV participated in behavioral tests and morphological analysis; DSMN performed the viral injections and helped with molecular and morphological analysis; ATC helped with molecular biology experiments; MI, MB, JB and MK performed the microarray analysis; NS, LP and JFO supervised the study, secured funding, analyzed the data and wrote the manuscript.
110 resilience of IP3R2 KO mice to the installation of an anxious-like phenotype induced by stress exposure, which suggests an important therapeutic potential. Materials and Methods Animals 3 to 5-month old WT and IP3R2 KO mice maintained on C57BL/6J background (Guerra-Gomes et al. 2018) were bred according to the guidelines for care and handling of laboratory animals ( ad libitum access to food and water in their home cages; lights maintained on a 12h light/dark cycle; 22 ± 1°C, 55% humidity). These mice were kindly supplied by Prof. Alfonso Araque (U. Minnesota, USA) (Navarrete et al. 2012), under agreement with Prof. Ju Chen (U.C. San Diego, USA) (Li et al. 2005). IP3R2 KO and WT littermate mice were obtained from crossing heterozygous animals, and their genotype was confirmed by polymerase chain reaction (PCR). Age and gender-matched sets of WT and IP3R2KO mice were divided into two experimental groups (n=7 WT + 6 IP3R2 KO per group): control (gently handled for 6 weeks) and uCMS-exposed mice. All experimental procedures were performed in accordance with the guidelines described in Directive 2010/63/EU and were approved by the local ethical committee (SECVS 075/2015) and national authority for animal experimentation (DGAV 17469/2012). unpredictable Chronic Mild Stress (uCMS) protocol Male WT and IP3R2 KO mice were exposed to a validated unpredictable Chronic Mild Stress (uCMS) protocol as previously described (Bessa et al. 2009a; Mateus-Pinheiro et al. 2013). Briefly, the uCMS paradigm consisted of a random exposure to mild stressors in an unpredictable fashion for six weeks. The applied stressors were as follows: shake – mice placed in an orbital shaker for 1 h at 150 rpm; strobe lights – light flashes exposure for 2 h; tilted cage – homecage tilted in a 45° angle overnight; damp bed – homecage with damp sawdust overnight; confinement – mice were limited to a restricted space for 1 h; overcrowding and noise – mice were placed in a restricted space and exposed to noise for 1 h; inverted lights – reversed light/dark cycle during the weekend. Biometric parameters assessment: body weight and coat state Stress efficacy was controlled through weekly body weight and fur state (coat score) monitoring. This assessment was performed on the same weekday and same time period. Coat score evaluation is a readout for an anhedonic phenotype, as previously described (Nollet et al. 2013). Stressed mice display a worse coat state as compared to control non-stressed animals. The coat score was attributed according
111 to the fur state in seven different body areas: head, neck, dorsal coat, ventral coat, tail, forepaws, and hindpaws. This classification was 0 (bad) if the fur was unkempt and 1 (good) if the coat presented a wellgroomed state. The sum of the scores for the seven body parts reached a maximal score of seven, whenever mice were motivated for self-care. Since some stressors, as damp bed and confinement, may influence coat state we assessed this biometric parameter in a period distanced from the exposure to these stressors. Corticosterone levels We assessed corticosterone levels in WT and IP3R2 KO mice from both groups. This assessment was performed in the blood serum using a Corticosterone ELISA kit (ab108821, Abcam, UK), following the manufacturer’s instructions. Blood sampling by tail venipuncture was performed within a maximum of 120 s after the mouse was removed from its homecage. Next, each sample was centrifuged at 13.000 rpm during 5 min and the supernatant (blood serum) was collected and stored at -80°C. Blood serum corticosterone levels were assessed in the last week of the uCMS protocol at the diurnal nadir (8-9 a.m) and diurnal zenith (8-9 p.m). Behavioral analysis After chronic stress exposure for 6 weeks, we performed behavioral experiments to assess emotional and cognitive phenotypes. Emotional states are the result of both physical and psychological alterations in response to internal or external stimuli that influence behavior (Fox 2008). Anxious-like behavior was assessed by the Elevated Plus Maze, depressive-like behavior was tested in the Tail Suspension Test and cognition was evaluated in the Novel Object Recognition test. Elevated-plus maze The Elevated-plus maze (EPM) is a gold standard method to assess anxious-like behavior in rodents (Belzung and Griebel 2001; Bourin et al. 2007). EPM test relies on propensity of mice to explore secure enclosed spaces and an intrinsic fear of heights/open zones (Walf and Frye 2007; Arabo et al. 2014). WT and IP3R2 KO mice, both control and exposed to the uCMS paradigm, were placed in the hub of a plus-like apparatus elevated 72.4 cm above the floor (ENV560; Med Associates Inc, Vermont, USA), containing two open arms (50.8 cm x 10.2 cm) and two closed arms (50.8 cm x 10.2 cm x 40.6 cm). Mice were let freely explore the maze for 5 min. Time spent in each arm and the number of entries were recorded and analyzed using Ethovision XT 13 software (Noldus, Netherlands).
112 Tail Suspension Test The Tail Suspension Test (TST) evaluates learned helplessness. In this test, a mouse suspended by its tail tends to develop an immobile state in a situation of behavior despair. During the total duration of the test (6 min), periods of mobility vs. immobility are analyzed (Steru et al. 1985). WT and IP3R2 KO mice (control and uCMS-exposed) were suspended by the tip of their tails using adhesive tape (approximately 80 cm above the floor). The activity of each mouse was recorded using a video camera. Mobility vs. immobility periods were analyzed using Ethovision XT 13 software (Noldus, Netherlands). Since this test evaluates learned helplessness, mice had to learn that there is no possible escape. Thus, only the final 4 min of testing were considered for analysis, for each animal. Regarding the latency to immobility, one animal was not included in the analysis due to an initial failure of the video-recording system. Novel Object Recognition Test The Novel Object Recognition Test (NOR) test was performed in white boxes (30 x 30 x 30 cm) under dim white-light room illumination. Boxes contained two visual cues (two stripes) on one side of the arena wall, as spatial reference. During the first three days, mice were let to explore the empty boxes during 20 min for habituation. The next day, a training session consisted of exposing mice to two equal objects during 10 min. After one hour, spatial recognition memory was assessed. One of the familiar objects was displaced to the opposite side of the box and mice were allowed to explore it during 10 min. After 24 h mice were tested for long-term memory. A novel object replaced a previously explored familiar object and mice were allowed to explore it during 10 min. The arena was cleaned between trials with 10% ethanol. This test was recorded with a video camera and analyzed using Ethovision XT 13 software (Noldus, Netherlands). Total object exploration time was extracted to calculate the percentage of exploration for the novel or displaced objects. Two mice were removed from the analysis because they did not perform the test (no exploration of the objects). Statistical analysis GraphPad Prism 7 (GraphPad Software Inc., USA) was used to assess normal distributions and further group comparisons. Parametric tests were applied to continuous variables following a Gaussian distribution (Kolmogorov-Smirnov), while non-parametric tests were used for discrete variables. Two-way analysis of variance (ANOVA) with Sidak post hoc test was applied to body weight gain, corticosterone levels, EPM, TST, and NOR parameters, considering either factors: control vs. uCMS or genotype. For coat score, a discrete variable, Kruskal-Wallis test was used, followed by Dunn's multiple comparisons
113 test. Data are expressed as mean ± SEM, except for coat score, which is represented as a scatter dot plot and median. The results were considered statistically significant for a p < 0.05. Results IP3R2 KO mice display resilience to uCMS exposure induced alterations in body weight and coat score WT and IP3R2 KO uCMS mice were exposed to several stressors in a random and unpredictable fashion (Figure 4.1A). To evaluate the efficiency of the uCMS protocol, we performed a weekly monitoring of body weight gain and coat score for all tested groups. In the last week of uCMS exposure, serum corticosterone levels were measured. Afterwards, a behavioral assessment covering distinct behavioral dimensions (anxiety-like behavior, depressive-like behavior, and cognition) was performed (Figure 4.1B). The biometrical assessment showed a significant impact of stress on body weight gain at week 6 of the uCMS protocol (Figure 4.2A; F 1, 22 = 24.02, p < 0.0001). Moreover, an interaction between stress exposure and genotype was found ( F 1, 22=4.464, p = 0.046), which was related to a smaller body weight gain in uCMS-exposed mice. Surprisingly, Sidak’s post hoc comparison revealed that WT-uCMS mice significantly lost weight after 6 weeks of uCMS exposure ( t 22 = 5.162, p < 0.0001), while IP3R2 KO-uCMS mice did not. Regarding self-care, our results showed that exposure to stress significantly impacted WT-uCMS coat Figure 4.1 – Experimental design for the uCMS protocol and behavioral characterization. (A) Schematic example of weekly stressors exposure. Once a week, the body weight was measured, and coat score was assessed. Blood collection was performed at the 6th week for corticosterone levels determination. (B) The behavioral assessment was performed in the weeks following uCMS exposure and covered three dimensions: anxious-like behavior tested in Elevated Plus Maze (EPM), mood assessed in the Tail Suspension Test (TST) and cognition evaluated in the Novel Object Recognition (NOR).
114 state, at week 6, (Figure 4.2B, t 22 = 3.666, p = 0.003), but the same was not observed for IP3R2 KO-uCMS mice. Serum corticosterone measurements, a proxy of hypothalamic-pituitary-adrenal (HPA) function, revealed a significant effect of uCMS exposure during 6 weeks (Figure 4.2C, F 1, 22 = 7.465, p = 0.012). However, post hoc analysis shows no differences between groups, pointing to a similar stress impact on corticosterone levels in WT and IP3R2 KO-uCMS mice. Altogether, our results show that both WT and IP3R2 KO mice exposed to uCMS present a dysregulation in corticosterone levels, reflecting the efficacy of chronic stress protocols. In sum, mice that lack IP3-dependent astrocytic Ca2+ signaling display a remarkable resilience to the establishment of the depressive phenotype, as shown by body weight maintenance and an increased self-care, even in the presence of a HPA-axis dysregulation. Figure 4.2 – Exposure to uCMS induces reduction of body weight and deteriorates self-care in WT, but not IP3R2 KO mice. (A) Body weight gain in WT and IP3R2 KO mice from control and uCMS-exposed groups after 6 weeks of uCMS exposure. (B) Coat score assessment in WT and IP3R2 KO mice from control and uCMS-exposed groups after 6 weeks of uCMS exposure. (C) Night/day ratio of corticosterone levels in the serum of WT and IP3R2 KO mice from control and uCMS-exposed groups after 6 weeks of uCMS exposure. WT mice represented in black and IP3R2 KO mice in red dots and bars. CTR group is represented in filled and uCMS group is represented in empty dots/bars. Body weight gain and corticosterone levels are plotted as mean ± SEM. Coat score is plotted as median. n=6–7 per group. * p < 0.05, ** p < 0.01, **** p < 0.0001 AB C CTR uCMS -2 0 2 4 6 **** Coat score
115 Lack of IP3R2-dependent astrocytic calcium signaling prevents an anxious-like phenotype after uCMS exposure Stress is associated with complex interactions between distinct behavioral domains as anxiety, mood, and cognition (Bessa et al. 2009b). We assessed anxiety-like behavior in the EPM (Figure 4.3A). Our results show a similar performance for both groups regarding the percentage of entries in open arms. However, a significant effect of genotype was found in the percentage of time spent in open arms ( F 1, 22 = 7.204, p = 0.014), which may be accounted by a pronounced decrease in the percentage of time spent in open arms observed in WT-uCMS exposed animals as compared with WT-CTR littermates ( t 22 = 2.747, p = 0.023). On the contrary, IP3R2 KO-uCMS mice spent a similar time in open arms when compared with their IP3R2 KO-CTR counterparts. Next, we performed a behavioral assessment for a depressive-like phenotype in the TST (Figure 4.3B). In this test, we analyzed the latency to immobility and the percentage of immobility as a measure of learned helplessness. We observed no differences between groups neither for latency to immobility, nor for immobility percentage. These results indicate that the Figure 4.3 – Assessment of anxiousand depressive-like behavior in WT and IP3R2 KO mice exposed to uCMS and respective non-stressed controls. (A) Elevated Plus Maze (EPM) test shows the percentage of entries and time spent in the open arms of the maze; (B) Tail Suspension Test (TST) evaluates latency to immobility and the percentage of time immobile in the last 4 min of testing. WT mice are represented in black bars and IP3R2 KO mice in red bars. CTR mice are represented in filled bars; uCMS-exposed mice are represented in empty bars. n=6–7 per group. Data plotted as mean ± SEM, * p < 0.05. Representative images obtained from https://neurology.mhmedical.com/data/books/1049/kan_ch63_f010.png (EPM) and https://www.creativebiolabs.com/drug-discovery/therapeutics/tail-suspension-test.htm (TST).
116 protocol used for uCMS exposure during 6 weeks was unable to induce a depressive-like state in WT, and thus we cannot conclude any effect on this dimension regarding IP3R2 KO mice. Next, we aimed at evaluating the impact of uCMS protocol in the cognitive dimension. For that, we performed the NOR test to assess spatial recognition (Figure S 4.1A) and long-term memory (Figure S 4.1B). Spatial recognition was evaluated by displacing one of the familiar objects, presented 1 hbefore, to the opposite side of the box. We found no differences in the performance of this task between WT and IP3R2 KO mice from both groups. In accordance, when a familiar object was replaced by a novel one 24 h later both groups performed similarly. In summary, our results show that uCMS exposure was able to induce an anxious-like phenotype in WT mice, but not in IP3R2 KO mice. However, exposing mice to an uCMS protocol during 6 weeks was unable to affect or indicate genotype-related effects in depressive-like behavior and cognition, at least for the behavioral paradigms tested. Discussion In this work, we addressed the influence of IP3R2-dependent astrocytic Ca2+ signaling in the context of depression. Growing evidence has shown that astrocytes undergo morphological and cell number alterations in depressive patients and in animal models of depression (Gosselin et al. 2009; Rajkowska and Stockmeier 2013; Rial et al. 2015; Wang et al. 2017). Astrocytes are actively involved in the modulation of synaptic function in brain regions affected by depression. While previous observations have suggested roles for astrocytic Ca2+ signaling in neurological and psychiatric diseases (Nedergaard et al. 2010), it is still missing to what extent astrocytic Ca2+ signals contribute to the installation and/or propagation of such pathologies (Shigetomi et al. 2016). Interestingly, recent studies reported a beneficial effect of IP3R2 genetic deletion in cases of brain insult or neurodegeneration (Li et al. 2015; Reichenbach et al. 2018). These results, together with our observations of cognitive enhancement in adulthood (Chapter 3), prompted us to test a putative protective role of IP3R2-dependent Ca2+ signaling in a chronic stress model that mimics anxious-like and depressive-like behavior of humans. Firstly, we applied an already validated protocol that mimics the biological correlates of depression – the uCMS protocol (Bessa et al. 2009a; Surget et al. 2011; Nollet et al. 2013; Gosselin et al. 2017). Our results indicate that we were able to induce a stress-response in WT mice on a C57BL/6J background, characterized by alterations in the anxious-state of these animals. However, the exposure of IP3R2 KO mice to uCMS revealed a surprising resilience to the deleterious effects of chronic stress. More
117 specifically, this resilience consisted of less body weight loss, more self-care behavior and reduced anxious-like phenotype, as observed in non-stressed IP3R2 KO mice. The available literature concerning the emotional consequences of the IP3R2 gene deletion are rather controversial. A previous study reported that control IP3R2 KO mice displayed a depressive-like behavior. The authors claim that this phenotype was due to decreased levels of astrocytic ATP release (Cao et al. 2013). Although we have not addressed the levels of gliotransmitters release from IP3R2 KO astrocytes, we observed that IP3R2 KO mice on a C57BL/6J background do not display a depressive-like phenotype (Chapter 3). In concomitance, the conditional deletion of IP3R2 specifically in astrocytes did not induce a depressive phenotype (Petravicz et al. 2014). In their study, Cao et al. (2013) employed several mouse models generated under distinct backgrounds. Therefore, the generation of mouse models with consistent genetic background and cellular specificity might help to clarify these observations (Oliveira et al. 2015). Another study reported that IP3R2 KO mice display increased learned helplessness after exposure to a chronic restraint stress (CRS) protocol (Monai et al. 2016). The CRS protocol consists of a repeated exposure to the same stressor. Protocols related to this paradigm display a high variability concerning its duration and extension (Moreira et al. 2016), which is critical for the impact of chronic stress. The CRS protocol might lead to distinct readouts as compared with the uCMS protocol. Beery and Kaufer (2015) discussed that stress response is influenced by a myriad of factors such as the time, nature and severity of the stressor. The protocol applied by Monai and co-workers (2016) subjected mice to an extensive period out of their homecages with no access to food or water, which is much more aggressive for mice. We believe that, by employing unpredictability of the presentation of distinct psychological and physical stressors we are testing a condition closer to daily life stress faced by humans. Body weight variations and coat state score are commonly used measures of stress efficacy in rodents (Nollet et al. 2013). We observed that, after 6 weeks of stress exposure, WT mice presented a significant body weight loss and coat state deterioration. Contrarily, IP3R2 KO-uCMS mice presented resilience to body weight variations and a better coat state. This result should not be due to metabolic alterations occurring in IP3R2 KO mice that could influence its body weight a priori , since our group and other studies confirmed that healthy IP3R2 KO mice gain weight similarly to WT littermates (Li et al. 2015; Pinto-Duarte et al. 2019). Body weight analysis should be considered as a measure for uCMS effectiveness, and not as a predictor of a certain behavioral response to it. For instance, Tau-KO mice display a reduction in body weight similar to the WT when exposed to stress. Even so, these mice display a protected response to stress in several behavioral domains, while WT mice reveal the deleterious stress effects (Lopes et al. 2016).
118 A stressful event, either with low or high intensity, influences self-grooming behavior in rodents (Kalueff and Tuohimaa 2005). Our results demonstrate that IP3R2 KO mice still display an interest for taking care of its fur state, whether WT littermates exposed to uCMS do not. The decreased coat score in WT mice goes in line with previous descriptions of a significant deterioration of the fur state after uCMS exposure (Ibarguen-Vargas et al. 2008). Here, we observed that interfering with IP3R2-dependent Ca2+ signaling prevented the deterioration of the coat state after uCMS exposure. A similar observation was found in stressed mice treated with an antidepressant, fluoxetine, which reversed the uCMS detrimental effect on coat state (Nollet et al. 2013). Chronic stress may affect distinct behavioral domains and dysregulate the HPA axis (de Kloet et al. 2005). As a result, disturbed corticosterone (CORT) levels are observed in mouse and rat models of depression (Monteiro et al. 2015; Patricio et al. 2015; Silva et al. 2018). Our results show that both genotypes presented a disruption in the healthy night/day ratio of CORT production in response to uCMS. These evidences suggest that the observed resilience to stress in IP3R2 KO mice is not due to preservation of the HPA axis rather it should appear downstream. CORT could act, at least partilly via astrocyte activation, and by silencing the global astrocyte Ca2+ signaling we could be blocking the deleterious pathways. Indeed, several studies showed that astrocytes express functional glucocorticoid receptors and might be key players in the development of the anxious-like phenotype (Tertil et al. 2018). Exposure to the 6-week uCMS protocol influenced anxious-like behavior in WT-uCMS mice, but IP3R2 KOuCMS mice were able to cope with it. Specifically, WT-uCMS mice spent less time in the open arms of the EPM, as compared with their control counterparts. Contrarily, IP3R2 KO mice exposed to uCMS display a similar exploration time of the open arms as compared to control IP3R2 KO mice. Different mouse models lacking IP3R2-dependent signaling were reported by several labs to be devoided of a basal anxiouslike phenotype (Tanaka et al. 2013; Petravicz et al. 2014; Pinto-Duarte et al. 2019). Therefore, the observed prevention of an uCMS-triggered anxious-like phenotype should be related with an IP3R2dependent response to stress. Despite having found a resilient response to anxiety-like behavior in IP3R2 KO mice, we were unable to observe a depressive-like phenotype or cognitive impairments in stressed mice from both genotypes. Depressive-like behavior was evaluated by TST. Both Forced Swimming Test and TST were considered pivotal tests to assess the validity of novel antidepressants. However, the relevance of immobility time in the context of depression is still under debate (Cryan and Holmes 2005). Therefore, coat state assessment emerged as a good measure for depressive-like behavior, and further demonstrated a
119 reversal effect following antidepressant treatment in uCMS mice (Santarelli et al. 2003; Nollet et al. 2013). In the future, this evaluation should be combined with a behavioral paradigm to assess anhedonia, such as the sucrose preference test, since both behavioral domains were shown to be correlated in depression (Bessa et al. 2009b). Regarding cognition, our results suggest that the chronic stress exposure failed to induce a phenotype of cognitive decline (Monteiro et al. 2015). Further studies, using different chronic stress protocols should be used to assess the relevance of astrocytic Ca2+-dependent signiling to support cognitive function. Overall, our results show that exposure to uCMS for 6 weeks was able to induce physiological and behavioral changes in WT mice. Surprisingly, IP3R2 KO-uCMS mice show less anxious-like behavior, maintain their weight and display more self-care, similarly to their non-stressed counterparts. This behavioral screening gathered relevant information on the role of astrocytic Ca2+ signaling in the function of healthy and stressed animals. These findings suggest that IP3R2-dependent Ca2+ signaling is required for the cellular mechanisms underlying the depressive pathology and that identifying key modulators for astrocytic function may afford for new therapeutic opportunities in the context of depression. Author Contributions SGG designed the study, performed and analyzed the experiments and wrote the manuscript; ARMS designed the study and performed the experiments; NDA and ELC helped with the uCMS protocol and behavioral tests; NS, LP and JFO supervised the study, secured funding, analyzed the data and wrote the manuscript.
126 (1) environments, as previously described (Sardinha et al. 2017). The maze was composed by three equal arms (33.2 L × 7 W × 15 cm H), made of white Plexiglas. To increase spatial recognition and navigation, the end of each arm contained a visual cue. Each mouse was initially placed at the end of the Start (S) arm and it was allowed to freely explore this arm and an additional arm (Familiar arm, F) for 5 min. In the second trial, a divider was removed allowing the exploration of a novel (N) arm, and each mouse was allowed to explore the three arms for 2 min for memory retrieval. The test was performed in dim light conditions and the maze was cleaned with 10% ethanol between subjects. All trials were acquired and analyzed using a video-tracking system (Videotrack; Viewpoint) and the EthoVision XT 12 software (Noldus, Netherlands). Data was expressed as a discrimination index (D.I.) of time and distance, calculated for the distal third of each arm using the following equation (1): 𝐷. 𝐼. = 𝑁𝑜𝑣𝑒𝑙 − 𝑆𝑡𝑎𝑟𝑡 + 𝐹𝑎𝑚𝑖𝑙𝑖𝑎𝑟 2 𝑁𝑜𝑣𝑒𝑙 + 𝑆𝑡𝑎𝑟𝑡 + 𝐹𝑎𝑚𝑖𝑙𝑖𝑎𝑟 2 A positive D.I. indicates preference for the novel arm, meaning that mice retained a memory of the arms previously explored (start and familiar), and therefore display a better spatial recognition performance. Tissue Processing and Immunohistochemical Analysis Mice were anesthetized with a mixture of ketamine (75 mg/kg, i.p.; Imalgene 1000, Merial, United States) and medetomidine (1 mg/kg, i.p.; Dorbene Vet, Pfizer, United States), and transcardially perfused with 0.9% saline. Brains were carefully removed, fixed overnight with 4% paraformaldehyde and immunofluorescence experiments were performed in cryostat coronal brain sections (20 µm thick). Sections were incubated overnight with the primary antibodies: rabbit anti-S100β (1:200, DakoCytomation; AB_2315306) and rabbit anti-NeuN (1:100, Cell Signaling Technology; AB_2651140). In the next day, incubation with the secondary antibody Alexa Fluor® 594 donkey anti-rabbit (1:1000, Thermo Fisher Scientific; AB_2556543) was carried out. Images were acquired in an Olympus Fluoview FV1000 confocal microscope (Olympus, Hamburg, Germany), and the number of S100β+ and NeuN+ cells was calculated using the ImageJ plugin – “Cell counter1”. Three Dimensional-Reconstruction of mPFC Layer V Pyramidal Neurons Three dimensional (3D) dendritic morphology was assessed in Golgi-Cox stained material as previously described (Lima et al. 2014). mPFC layer V pyramidal neurons were analyzed for the following dendritic 1 http://fiji.sc/Fiji
127 features: length, arborization, Sholl analysis and spine number and classification. Briefly, at least five neurons were analyzed for each animal by using a motorized microscope controlled by the Neurolucida software (MBF Bioscience, United States) under 100× magnification. Dendritic spine densities were assessed in randomly selected dendritic segments of 30 µm, in the proximal and distal portions of the apical dendrite and in the proximal portion of the basal dendrite. Moreover, spines were classified into four categories: thin, mushroom, thick, and ramified. The extraction of data for both reconstructed neurons and spines was performed by using NeuroExplorer software (MBF Bioscience, United States). Statistical Analysis Statistical analysis was performed using the GraphPad Prism 6.01 (GraphPad Software Inc., United States). All data analyzed passed the D’Agostino and Pearson normality test for Gaussian distributions. Two-way analysis of variance (ANOVA) with Sidak post hoc test was applied to analyze the performance in the Y-maze 2TPR, cell densities, neuronal length, and endings, considering either factor: genotype or age. Two-way ANOVA with Tukey’s multiple comparisons test was used to analyze Sholl analysis data for neuronal 3D reconstructions. Data are presented throughout the manuscript as mean ± SEM (Standard Error of the Mean) and results were considered significant for p < 0.05. Results Lack of IP3R2-Dependent Astrocytic Calcium Prevents Age-Related Cognitive Decline We tested the performance of both WT and IP3R2 KO mice (adult and aged) in a PFC-dependent task. We performed the 2TPR task which evaluates spatial recognition memory in mice, based on their natural drive to explore novelty (Figure 5.1). This test has the advantage of being physically less demanding, and therefore more suitable to assess cognition in aged mice (Pistell et al. 2012; Dolgin 2013). Our results show that aged WT mice display a deficit in recognition memory, since they retain less memory of the familiar arms and fail to discriminate the novel arm when compared to their adult WT littermates (Figure 5.1A,B; Sidak post hoc test; p < 0.05). Surprisingly, this deficit in spatial recognition memory is not observed in aged IP3R2 KO mice, as those animals explored the novel arm longer, similarly to their adult counterparts (Figure 5.1A,B). In accordance, aged WT mice walk less distance in the novel area than their adult WT littermates (Figure 5.1A,C; Sidak post hoc test; p < 0.05), a deficit not observed in aged IP3R2 KO mice. The Sidak post hoc comparison between genotypes discarded any significant difference for both measures (time or distance), excluding an effect of IP3R2 KO in the performance between adult or aged mice in this experimental setup. Importantly, tested mice equally explored the maze during the
128 task, as given by their total distance traveled (Figure 5.1D) and number of arm entries (Figure 5.1E), regardless of their genotype or age, hence excluding any age-related loss of their natural exploratory drive. Together, these results point out an unpredicted maintenance of PFC-dependent cognitive performance in aged mice that lack IP3-dependent astrocytic Ca2+ signaling. Aging Leads to a Decrease in Neuronal but Not Glial Densities in IP3R2 KO Mice To correlate the observed behavioral phenotype with possible alterations in the populations of neurons and astrocytes, we estimated the density of cells, respectively, stained for NeuN+ and S100β+ markers. NeuN is a well-known marker for post-mitotic neurons (Mullen et al. 1992), whereas S100β is a recognized marker for astrocytes (Wang and Bordey 2008). We assessed these cellular densities in layer V of the medial PFC (Figure 5.2A), which is the main prefrontal output to other cortical and subcortical regions involved in cognitive behavior (Opris and Casanova 2014; Naka and Adesnik 2016), and receives a main hippocampal input involved in episodic memory with a strong spatial component (Hoover and Vertes 2007; Barker et al. 2017) required for the performance in the 2TPR task. S100β+ cells were Figure 5.1 - Lack of IP3R2-dependent astrocytic calcium prevents age-related cognitive decline in a PFC-dependent task. (A) Representative heatmaps of cumulative time exploration of start (S), familiar (F), and novel (N) arms of the Y-maze for WT (left, black) and IP3R2 KO (right, red) mice (warm colors, more time; cold colors, less time); Discrimination index (D.I.) for time (B) and distance (C), representative of the spatial preference for the novel arm in the retrieval trial. (D) Total distance traveled in the maze; (E) Total number of entries in the three arms. Adult mice are represented in filled bars; aged mice are represented in unfilled bars. n= 9–13 per group, twoway ANOVA, Sidak’s multiple comparisons test; data plotted as mean ± SEM. ∗ p < 0.05.
129 previously shown to colocalize with IP3R2 in the rodent brain (Sharp et al. 1999; Takata et al. 2011; Li et al. 2015). Our analysis indicated an effect of aging on NeuN+ cell density (Figure 5.2B,D; age: F 1,33 = 6.192, p = 0.018), which may be accounted by a pronounced decrease of NeuN+ cells observed in aged IP3R2 KO mice when compared with their aged WT littermates (Figure 5.2D; Sidak post hoc test; p < 0.05). Moreover, this reduction in the number of neuronal cells in IP3R2 KO mice is also significantly different from their adult IP3R2 KO mice (Sidak post hoc test; p < 0.01), that maintain similar cellular densities to WT mice. On the contrary, we found an overall increase of S100β+ astrocytes in IP3R2 KO mice (Figure 5.2C,E; genotype: F 1,40 = 16.19, p < 0.05). Contrarily to what we observed for neurons, we found an increase in S100β+ cells in aged IP3R2 KO when compared to aged WT mice (Sidak post hoc test; p < 0.01). Altogether, these results suggest that lack of age-related cognitive impairment in the absence of IP3R2-mediated Ca2+ signaling in astrocytes, correlates with a different neuron-to-glia ratio in the PFC which could ultimately result either in different neuron-glial circuits, or functional neuron-glia interactions or both. Figure 5.2 - Aged IP3R2 KO mice display decreased NeuN+ neuron densities and increased S100β+ astrocyte densities. (A) Scheme of the prelimbic subregion of the medial prefrontal cortex (mPFC, blue) and respective layers I, II/III, and V. mPFC layer V (white dash line) corresponds to the area considered for cell counting. (B) Representative mPFC layer V confocal images of DAPI/NeuN labeling for WT and IP3R2 KO mice (adult and aged). (C)
130 Representative mPFC layer V confocal images of DAPI/S100β labeling for WT and IP3R2 KO mice (adult and aged); (A–C) scale bars = 200 µm. (D,E) Densities of NeuN+ (D) and S100β+ (E) cells in mPFC layer V of WT and IP3R2 KO mice (adult and aged; n=8–15 images per group; two-way ANOVA, Sidak’s multiple comparisons test). WT (adults and aged) mice are represented in filled/unfilled black bars, whilst IP3R2 KO (adults and aged) are represented in filled/unfilled red bars. Data plotted as mean ± SEM. # p < 0.05 and ## p < 0.01 between genotypes and ∗∗ p < 0.01 between ages. Aging Leads to a Dendritic Refinement of mPFC Layer V Pyramidal Neurons in IP3R2 KO Mice An important component underpinning neuron-glial interaction, is the neuropil morphology to the extent that it may define structural and functional constraints of such interaction (Volterra et al. 2014; De Pitta et al. 2016). With this regard, we considered dendrite morphology for WT vs. IP3R2 KO and adult vs. aged mice (Figure 5.3A). Significantly, both age and lack of IP3R2 lead to a decrease in apical dendrite length of 3D-reconstructed Golgi-impregnated neurons (Figure 5.3B; age: F 1,42 = 6.340, p = 0.016; genotype: F 1,42 = 5.168, p = 0.028), which was accompanied by a reduction of apical endings in mice that lack IP3R2 ( F 1,42 = 6.259, p = 0.016). In line with neuronal density data, these effects were caused by the marked reductions of apical dendrites of layer V pyramidal neurons observed in aged IP3R2 KO mice. Specifically, aged IP3R2 KO pyramidal neurons display a shorter and less ramified morphology, as compared with aged WT mice (Sidak post hoc test; p < 0.05). Furthermore, when compared with their adult IP3R2 KO littermates, aged IP3R2 KO mice also revealed significantly reduced apical dendritic length (Sidak post hoc test; p < 0.05). These alterations are detailed by the Sholl analysis data, which revealed an overall effect of radius, genotype and an interaction between these two factors (radius: F 22,924 = 41.86, p < 0.0001; genotype: F 3,42 = 4.089, p = 0.012; interaction: F 66,924 = 1.821, p = 0.0001). Post hoc analysis showed that aged IP3R2 KO mice display fewer intersections, as compared with aged WT and their adult genotype matches (Tukey post hoc test; p < 0.05). The analysis of basal dendritic morphology in neurons from the same layer V revealed that although they seem to be also reduced in aged IP3R2 KO mice, this is only significant in the overall comparison by age and in the detailed Sholl analysis. Specifically, aging seems to lead to dendritic shortening (Figure 5.3C; F 1,42 = 7.230, p = 0.010). Moreover, the Sholl analysis revealed a significant effect of radius, genotype and an interaction between factors (radius: F 17,714 = 160.4, p < 0.0001; genotype: F 3,42 = 3.532, p = 0.023; interaction: F 51,714 = 2.055, p < 0.0001). Additional analysis showed a decreased number of intersections around 40–100 µm from the soma in aged IP3R2 KO, as compared with aged WT and adult IP3R2 KO mice (Tukey post hoc test; p < 0.05). Since dendritic remodeling frequently has consequences for spine stability, we further analyzed the implication of IP3R2-
131 dependent signaling to spine integrity upon aging. For that, we performed dendritic spine categorization at the apical proximal, apical distal, and basal segments of these neurons (Figure 5.3D,E). We observed the typical distribution of spines in the three dendritic segments, being the thin and mushroom types much more abundant than thick or ramified types: apical proximal ( F 3,99 = 386.1, p < 0.0001); apical distal ( F 3,99 = 312.6, p < 0.0001); basal ( F 3,99 = 250.9, p < 0.0001). Besides this expected spine distribution, aging led to an increase in the mushroom type specifically in the basal dendrites in mice of both genotypes, which was accompanied by a decrease in the densities of thick spines (Tukey post hoc test; p < 0.05). This observation was also described previously (Burke and Barnes 2006) and was independent of the lack of IP3R2. Figure 5.3 - Aging leads to a dendritic refinement of mPFC layer V pyramidal neurons in IP3R2 KO mice. (A) Representative 3D reconstructions of layer V mPFC neurons from WT and IP3R2 KO mice (adults or aged; scale bar = 100 µm). (B,C) Morphological parameters of apical (B) and basal (C) dendrites by assessing total dendritic length, number of endings and Sholl intersections (n=10–14 neurons per group; two-way ANOVA, Sidak’s multiple comparisons test). (D) Representative images of basal dendritic segments of each experimental group; scale bar = 5 µm; arrowheads, mushroom-type; arrow, thick-type. (E) Percentage of each spine type at proximal and distal segments of apical dendrites, and basal dendrites (n=6–14 neurons per group; two-way ANOVA, Tukey’s multiple
132 comparisons test). WT (adults and aged) mice are represented either by filled/unfilled black bars or black line/dashed line, while IP3R2 KO (adults and aged) are represented either by filled/unfilled red bars or red line/dashed line. Data plotted as mean ± SEM. ∗ Denotes the effect of age; # denotes the effect of genotype. Sholl analysis: black ∗ ,refers to difference between WT; red ∗ , refers to difference between IP3R2 KO; #, ∗ p < 0.05; ∗∗ p < 0.01. Discussion In the present study we applied behavioral and morphological approaches to untangle the influence of IP3R2-dependent astrocytic Ca2+ in the aged PFC. We found memory impairment in aged WT mice, which is in accordance with several evidences pointing to a cognitive decline during aging (Rajah and D'Esposito 2005; Grady 2008; Weber et al. 2015). Surprisingly, aged mice that lack astrocyte Ca2+ elevations (IP3R2 KO) perform similarly to WT mice in the same PFC-dependent task, suggesting that astrocytes are involved in the age-related cognitive decline. Regarding the behavioral observations, the cognitive conservation of aged IP3R2 KO mice is in line with previous studies using the same mouse model that show a putative role for IP3R2-dependent Ca2+ signaling in neuroprotection and behavior conservation after brain damage (Li et al. 2015; Rakers and Petzold 2017). More importantly, a recent study shows that the deletion of IP3R2 in a AD model (typically associated with cognitive decline) leads to the retention of spatial learning and memory (Reichenbach et al. 2018). This suggests that astrocyte activity mediated by reticular Ca2+ elevations may influence the network structure and function along the aging process. Indeed, Boisvert et al. (2018) demonstrated that in spite of maintaining homeostatic and neurotransmission-regulating genes, aged astrocytes in the cortex partially resemble reactive astrocytes creating an environment permissive to synapse elimination and neuronal damage, possibly contributing to aging-associated cognitive decline. The available literature appears to agree that normal aging does not lead to major neuronal loss in most cortical regions (Burke and Barnes 2006; Bishop et al. 2010). Besides, the neuron-glia signaling seems to remain conserved during lifespan (Gomez-Gonzalo et al. 2017). Our data are in agreement with these evidences, since both neuronal (NeuN+) and astrocyte (S100β+) densities remain unchanged in aged WT mice. Curiously, aged IP3R2 KO mice – that retain the cognitive ability in a PFC-dependent task – display a reduction in neuronal densities, while astrocyte densities display an inverse tendency in the same region, when compared to the observations of us and others at the adult stage (Li et al. 2015). These findings may be linked since overexpression of astrocyte markers in aged mice has been associated with its detrimental effects to neuronal networks in several cortical areas, namely through astrogliosis (Lynch
133 et al. 2010; Orre et al. 2014; Rodriguez et al. 2014; Rogers et al. 2017; Boisvert et al. 2018). While healthy aging is linked with mild or subtle morphological changes, pathological aging is rather linked to drastic reductions of neuronal structures (Hof and Morrison 2004; Dickstein et al. 2007; Konsolaki and Skaliora 2015). In this work, aged IP3R2 KO mice display a marked reduction of dendritic tree complexity, namely in apical dendrites, which appears to be related with the retention of their cognitive abilities as they age. Although our data does not provide a causal link between structural changes (dendritic simplification and different cell densities) and “silenced” age-related astrocyte signals, we believe that the preserved cognitive performance is rather a result of both. These neuronal alterations may result from changes in activation and signaling of a larger number of astrocytes (similarly to a mild astrogliosis), additional glial involvement (i.e., microglia, that play an extensive role in brain aging) (Lynch et al. 2010; Lalo et al. 2014; Das and Svendsen 2015; Verkhratsky et al. 2016b), and homeostatic regulatory mechanisms at the neuron network level (Wefelmeyer et al. 2016). It is noteworthy that our data indicate a great degree of spine stability in aged mice, namely at apical dendrites. Spine distribution and number in the PFC is tightly related with cognitive performance and evolves along the aging process (Burke and Barnes 2006; Holtmaat and Svoboda 2009; Bloss et al. 2011). In basal dendrites, aging triggers a shift to the mushroom type, which refers to a need for synaptic stability and complexity in aged layer V synapses. Nevertheless, this effect was visible in tissue of both WT and IP3R2 KO mice, which suggests that spine remodeling may not play a relevant role in the observed cognitive maintenance. Conclusion In conclusion, the present study demonstrates that silencing the main source of global astrocytic Ca2+ leads to a conservation of the cognitive ability along aging. This cognitive resilience appears to be a product of gross neuronal structure refinement, together with an addition of astrocytes to the network. Whether other cell types and/or circuits might be involved in this cognitive maintenance is a matter that should be addressed in the future. Either way, disclosing further the astrocyte roles in the aging brain should be pivotal to provide alternative solutions to prevent or treat cognitive decline in humans. Author Contributions SGG, JV, DSMN, JC, VS, and IC designed, performed, and analyzed the experiments; composed the figures; and wrote the manuscript. NS, LP, and JO supervised the study, wrote the manuscript, and secured funding.
134 CHAPTER 6 General Discussion and Conclusions
135 Astrocytes respond to excitatory or inhibitory receptor activation with intracellular Ca2+ elevations, that may lead to downstream effects, such as release of gliotransmitters with impact in synaptic function. These Ca2+ elevations display complex and specific spatio-temporal properties that control synapses, circuits and behavior (Volterra et al. 2014; Guerra-Gomes et al. 2017). Since the emergence of the tripartite synapse concept (Araque et al. 1999), several efforts have been made in the field to develop tools to study the role of astrocytes in the control of higher brain functions, namely by assessing its Ca2+ dynamics. From the initial studies using Ca2+ dyes in cell culture (Cornell-Bell et al. 1990; Charles et al. 1991), to the development of genetically encoded Ca2+ indicators (Shigetomi et al. 2016) a long way has been traveled. However, there is still a longer way to go. For instance, it remains unexplored how different astrocytic Ca2+ signals in distinct brain areas could affect brain functioning. Moreover, the association of a diversity of Ca2+ sources with the astrocytic response to synaptic modulation is unknown. Until now, the contribution of a Gq-GPCR signaling pathway triggering Ca2+ elevations in astrocytes is widely recognized. The activation of this pathway results in the production of the IP3 molecule, which binds to the IP3Rs present in the ER and leads to Ca2+ release from this intracellular store. In the CNS, the IP3R2 was proven to be mainly expressed in astrocytes (Sharp et al. 1999; Hertle and Yeckel 2007; Takata et al. 2011; Zhang et al. 2014; Li et al. 2015). It corresponds to the major source of ER-dependent global astrocytic Ca2+ and mice lacking this receptor display an impairment in Ca2+ elevations mainly in the soma and main processes (Petravicz et al. 2008; Takata et al. 2011; Navarrete et al. 2012). In this thesis, our main goal was to explore the role of IP3R2-dependent Ca2+ signaling in cortico-limbic function and behavior. Our main findings are summarized in Figure 6.1. Figure 6.1 – Scheme depicting the behavior characterization performed in IP3R2 KO mice throughout the lifespan.
142 Aged mice lacking astrocytic reticular Ca2+ display preserved cognitive performance in a PFC-dependent task The differential cognitive phenotype observed in adult mice led us to assess the evolution of cognitive performance along healthy aging. This work was published in Frontiers in Cellular Neuroscience (GuerraGomes et al. 2018). We performed the Y-Maze 2TPR task, which has the advantage of being less physically demanding, and consequently more adjusted to evaluate cognitive function in aged mice (Pistell et al. 2012; Dolgin 2013). Similar to previous observations in aged humans and rodents (Rajah and D'Esposito 2005; Grady 2008; Weber et al. 2015), we found an impaired PFC-dependent cognitive performance in aged WT mice. In opposition, aged IP3R2 KO mice performed similarly to their adult counterparts, showing a surprising resilience to age-related cognitive decline in this test. This result goes in line with a recent study using an AD mouse model lacking IP3R2 (Appps1+/− × Ip3r2−/−), which displayed a protected cognitive performance in a spatial reference memory task (Reichenbach et al. 2018). This striking result led us to address morphological and cellular correlates of network function in the mPFC layer V. This layer contains the major excitatory neurons, responsible for cortical-cortical and cortical-subcortical projections (Tang et al. 2014). This cellular characterization provided evidences for conserved neuronal and astrocyte cell numbers in aged WT mice, as previously described (Burke and Barnes 2006; Bishop et al. 2010; Gomez-Gonzalo et al. 2017). Such result is supported by observations, in both humans and rodents, describing that cognitive decline is more associated with synaptic changes rather than neuronal loss (Berchtold and Cotman 2009). We observed that adult IP3R2 KO mice have similar neuronal and astrocytic densities in mPFC layer V, as compared with WT littermates, as previously reported (Li et al. 2015). Absence of global astrocytic Ca2+ signaling leads to a decreased density of NeuN+ cells in aged mice, which is compensated by an increased number of S100β+ cells, and may account for the preserved cognitive performance observed in aged IP3R2 KO mice. In Chapter 3, we have performed a morphological characterization of CA1 pyramidal neurons in WT and IP3R2 KO mice. Here, we took advantage of the same strategy to evaluate mPFC layer V pyramidal neurons, in adult and aged mice. Similar to CA1 pyramidal neurons morphology, we observed no differences between adult WT and IP3R2 KO mPFC layer V neurons. In opposition, we found no alterations regarding the percentage of dendritic spines in WT and IP3R2 KO mice. This result is interesting since it points to a region-specific effect of IP3R2-dependent Ca2+ signaling in the mPFC and the hippocampus: while the former display no alterations, the latter has apical dendrites with a higher percentage of immature spines. The present study revealed that aged IP3R2 KO mice presented a neuronal dendritic refinement in the PFC. This result is in accordance with a recent work from our lab, stating a better
143 performance in aged rats presenting a shorter dendritic arborization. In fact, animals that were classified as “bad performers” in PFCand hippocampal-dependent tasks displayed increased dendritic arborization and impaired autophagy (Mota et al. 2019). Curiously, a link between astrocytic function and autophagy was described in cultures of cerebellar cells from a mouse model of a neurodegenerative disease (Kanner et al. 2018). Thus, it would be interesting to look for autophagy markers in the PFC of aged WT and IP3R2 KO, to correlate possible alterations with dendritic morphology and behavioral results. Taken together, our results demonstrated an IP3R2-related maintenance of PFC-dependent behavioral phenotype during aging, which is supported by altered densities of NeuN+ and S100β+ cells and a neuronal dendritic refinement in the PFC of aged IP3R2 KO mice. Considering the results of Chapter 3, FOXO1related mechanisms could also play a role in the morphology of the aged prefrontal circuits influencing synaptic coverage and signaling, with impact in behavior.
144 Conclusion and Future Perspectives The work developed during this PhD thesis demonstrates, for the first time, that interfering with the major source of global Ca2+ signaling in astrocytes does not influence postnatal development. Therefore, the IP3R2 KO mouse model is suitable to study astrocyte involvement in the behavioral response, in all developmental stages and adulthood. Our results highlight an important role of IP3R2-dependent Ca2+ signaling in astrocytes in the regulation of cognitive behavior in the adult and aged brain. In the adult age, we found that IP3R2 KO mice display an enhanced cognitive performance at hippocampal-dependent tasks. Our results provided the first evidence of a transcription factor, Foxo1 , which is highly enriched in astrocytes and whose activity seems to be controlled by global Ca2+ signaling in astrocytes. Foxo1 is regulating the expression of relevant astrocytic-related genes in the hippocampus - Sdc2 and Ezr - that have an important role in PAP morphology and spinogenesis (Heller and Rusakov 2015; Hu et al. 2016), which in turn should control synaptic plasticity phenomena to enhance cognitive performance. Moreover, we have provided evidence for the role of astrocytic IP3R2-dependent Ca2+ signaling in the stressed brain. Specifically, we observed that IP3R2 KO mice display a resilience to physical and anxious-like alterations, typically caused by chronic stress exposure. The disclosure of astrocytic molecular pathways involved in this resilience may add therapeutical advances to improve behavioral deficits related to depression. Finally, in healthy aging, we found that cognitive function dependent on another cortico-limbic area, the PFC, was preserved in mice lacking IP3R2-dependent Ca2+ signaling in astrocytes. These results open a new window in the search for therapeutic targets/mechanisms, specifically in astrocytes, to treat agedrelated cognitive loss. In this thesis, we showed that long-lasting, slower Ca2+ events that rely on reticular Ca2+ influence corticolimbic function and behavior. We performed a multifactorial assessment of its implications in brain functioning throgout life (from postnatal development to aging; health vs disease context). These results open several windows of research that should be further studied to target specific spatio-temporal dynamics of Ca2+ signaling in astrocytes: 1) This thesis work pointed Foxo1 and its regulated transcription genes ( Sdc2 and Ezr ) as potential astrocytic mediators to control cognitive performance in the hippocampus. Therefore, it is important to assess the expression levels of Foxo1 target genes, Sdc2 and Ezr , (and others identified in the STRING analysis, even if they are not astrocyte-specific) in astrocyte-enriched samples, not only in the adult mouse, but also in the aged mouse and in the chronically stressed
145 mouse. This complete analsyis will elucidate us regarding the therapeutic potential of global Ca2+- Foxo1 pathway modulation. 2) The molecular dissection of the global Ca2+- Foxo1 pathway modulation in these 3 conditions, should be correlated with a behavior analysis of mice that locally overexpress FOXO1 in the hippocampus or PFC, to confirm behavior recovery or protection. 3) We found that IP3R2 KO-dependent Ca2+ signaling regulates cognitive performance in hippocampal-dependent tasks. Therefore, it is pivotal to perform electrophysiological recordings to obtain a functional readout of cortico-limbic structures, namely related with hip-PFC regional activity and temporal synchronization, to better understand the link between molecular alterations and cognitve enhancement or resilience. 4) The IP3R2 KO mouse model has a similar reduction in global Ca2+ signaling as the recently generated viral tool to genetically drive the continuous extrusion of cytosolic Ca2+ in astrocytes (Yu et al. 2018). Therefore, this tool should be used to control Ca2+ signals locally in physiological processes to fully understand the implications of IP3R2-dependent Ca2+ signaling in distinct brain regions and under different levels of activation.
146 References Adamsky A, Kol A, Kreisel T, Doron A, Ozeri-Engelhard N, Melcer T, Refaeli R, Horn H, Regev L, Groysman M, London M, Goshen I. 2018. Astrocytic Activation Generates De Novo Neuronal Potentiation and Memory Enhancement. Cell 174:59-71 e14. Agarwal A, Wu PH, Hughes EG, Fukaya M, Tischfield MA, Langseth AJ, Wirtz D, Bergles DE. 2017. Transient Opening of the Mitochondrial Permeability Transition Pore Induces Microdomain Calcium Transients in Astrocyte Processes. Neuron 93:587-605 e7. Aguado F, Espinosa-Parrilla JF, Carmona MA, Soriano E. 2002. Neuronal activity regulates correlated network properties of spontaneous calcium transients in astrocytes in situ. J Neurosci 22:943044. Agulhon C, Boyt KM, Xie AX, Friocourt F, Roth BL, McCarthy KD. 2013. Modulation of the autonomic nervous system and behaviour by acute glial cell Gq protein-coupled receptor activation in vivo. J Physiol 591:5599-609. Agulhon C, Fiacco TA, McCarthy KD. 2010. Hippocampal shortand long-term plasticity are not modulated by astrocyte Ca2+ signaling. Science 327:1250-4. Agulhon C, Petravicz J, McMullen AB, Sweger EJ, Minton SK, Taves SR, Casper KB, Fiacco TA, McCarthy KD. 2008. What is the role of astrocyte calcium in neurophysiology? Neuron 59:932-46. Agulhon C, Sun MY, Murphy T, Myers T, Lauderdale K, Fiacco TA. 2012. Calcium Signaling and Gliotransmission in Normal vs. Reactive Astrocytes. Front Pharmacol 3:139. Allen NJ. 2014. Astrocyte regulation of synaptic behavior. Annu Rev Cell Dev Biol 30:439-63. Allen NJ, Barres BA. 2009. Neuroscience: Glia - more than just brain glue. Nature 457:675-7. Allen NJ, Lyons DA. 2018. Glia as architects of central nervous system formation and function. Science 362:181-185. Alvarez JI, Katayama T, Prat A. 2013. Glial influence on the blood brain barrier. Glia 61:1939-58. Alves ND, Correia JS, Patricio P, Mateus-Pinheiro A, Machado-Santos AR, Loureiro-Campos E, Morais M, Bessa JM, Sousa N, Pinto L. 2017. Adult hippocampal neuroplasticity triggers susceptibility to recurrent depression. Transl Psychiatry 7:e1058. Amorim L, Magalhaes R, Coelho A, Moreira PS, Portugal-Nunes C, Castanho TC, Marques P, Sousa N, Santos NC. 2018. Poor Sleep Quality Associates With Decreased Functional and Structural Brain Connectivity in Normative Aging: A MRI Multimodal Approach. Front Aging Neurosci 10:375. Arabo A, Potier C, Ollivier G, Lorivel T, Roy V. 2014. Temporal analysis of free exploration of an elevated plus-maze in mice. J Exp Psychol Anim Learn Cogn 40:457-66. Araque A. 2008. Astrocytes process synaptic information. Neuron Glia Biol 4:3-10. Araque A, Carmignoto G, Haydon PG, Oliet SH, Robitaille R, Volterra A. 2014. Gliotransmitters travel in time and space. Neuron 81:728-39. Araque A, Navarrete M. 2010. Glial cells in neuronal network function. Philos Trans R Soc Lond B Biol Sci 365:2375-81. Araque A, Parpura V, Sanzgiri RP, Haydon PG. 1999. Tripartite synapses: glia, the unacknowledged partner. Trends Neurosci 22:208-15. Bailey C, Barco A, Hawkins R, Kandel E. 2008. 4.02 - Molecular studies of learning and memory in Aplysia and the hippocampus: a comparative analysis of implicit and explicit memory storage. In: Byrne J, editor. Learning and memory: a comprehensive reference Oxford, UK: Elsevier Press. Baldessarini RJ, Lau WK, Sim J, Sum MY, Sim K. 2015. Duration of initial antidepressant treatment and subsequent relapse of major depression. J Clin Psychopharmacol 35:75-6. Banasr M, Chowdhury GM, Terwilliger R, Newton SS, Duman RS, Behar KL, Sanacora G. 2010. Glial pathology in an animal model of depression: reversal of stress-induced cellular, metabolic and behavioral deficits by the glutamate-modulating drug riluzole. Mol Psychiatry 15:501-11.
147 Banasr M, Duman RS. 2008. Glial loss in the prefrontal cortex is sufficient to induce depressive-like behaviors. Biol Psychiatry 64:863-70. Barker GR, Banks PJ, Scott H, Ralph GS, Mitrophanous KA, Wong LF, Bashir ZI, Uney JB, Warburton EC. 2017. Separate elements of episodic memory subserved by distinct hippocampal-prefrontal connections. Nat Neurosci 20:242-250. Barragan M, de Frias M, Iglesias-Serret D, Campas C, Castano E, Santidrian AF, Coll-Mulet L, Cosialls AM, Domingo A, Pons G, Gil J. 2006. Regulation of Akt/PKB by phosphatidylinositol 3-kinasedependent and -independent pathways in B-cell chronic lymphocytic leukemia cells: role of protein kinase C{beta}. J Leukoc Biol 80:1473-9. Bayraktar OA, Fuentealba LC, Alvarez-Buylla A, Rowitch DH. 2014. Astrocyte development and heterogeneity. Cold Spring Harb Perspect Biol 7:a020362. Bazargani N, Attwell D. 2016. Astrocyte calcium signaling: the third wave. Nat Neurosci 19:182-9. Bechtholt-Gompf AJ, Walther HV, Adams MA, Carlezon WA, Jr., Ongur D, Cohen BM. 2010. Blockade of astrocytic glutamate uptake in rats induces signs of anhedonia and impaired spatial memory. Neuropsychopharmacology 35:2049-59. Beery AK, Kaufer D. 2015. Stress, social behavior, and resilience: insights from rodents. Neurobiol Stress 1:116-127. Belzung C, Griebel G. 2001. Measuring normal and pathological anxiety-like behaviour in mice: a review. Behav Brain Res 125:141-9. Ben Haim L, Rowitch DH. 2017. Functional diversity of astrocytes in neural circuit regulation. Nat Rev Neurosci 18:31-41. Benediktsson AM, Marrs GS, Tu JC, Worley PF, Rothstein JD, Bergles DE, Dailey ME. 2012. Neuronal activity regulates glutamate transporter dynamics in developing astrocytes. Glia 60:175-88. Berchtold NC, Cotman CW. 2009. Normal and Pathological Aging: From Animals to Humans. In: Bizon JL, Woods A, editors. Animal Models of Human Cognitive Aging. Totowa, NJ: Humana Press. p 1-28. Bernardinelli Y, Muller D, Nikonenko I. 2014. Astrocyte-synapse structural plasticity. Neural Plast 2014:232105. Bessa JM, Ferreira D, Melo I, Marques F, Cerqueira JJ, Palha JA, Almeida OF, Sousa N. 2009a. The mood-improving actions of antidepressants do not depend on neurogenesis but are associated with neuronal remodeling. Mol Psychiatry 14:764-73, 739. Bessa JM, Mesquita AR, Oliveira M, Pego JM, Cerqueira JJ, Palha JA, Almeida OF, Sousa N. 2009b. A trans-dimensional approach to the behavioral aspects of depression. Front Behav Neurosci 3:1. Bibb JA, Mayford MR, Tsien JZ, Alberini CM. 2010. Cognition enhancement strategies. J Neurosci 30:14987-92. Biesecker KR, Srienc AI, Shimoda AM, Agarwal A, Bergles DE, Kofuji P, Newman EA. 2016. Glial Cell Calcium Signaling Mediates Capillary Regulation of Blood Flow in the Retina. J Neurosci 36:943545. Biggs WH, 3rd, Meisenhelder J, Hunter T, Cavenee WK, Arden KC. 1999. Protein kinase B/Akt-mediated phosphorylation promotes nuclear exclusion of the winged helix transcription factor FKHR1. Proc Natl Acad Sci U S A 96:7421-6. Bindocci E, Savtchouk I, Liaudet N, Becker D, Carriero G, Volterra A. 2017. Three-dimensional Ca2+ imaging advances understanding of astrocyte biology. Science 356. Bishop NA, Lu T, Yankner BA. 2010. Neural mechanisms of ageing and cognitive decline. Nature 464:529-35. Bissonette GB, Martins GJ, Franz TM, Harper ES, Schoenbaum G, Powell EM. 2008. Double dissociation of the effects of medial and orbital prefrontal cortical lesions on attentional and affective shifts in mice. J Neurosci 28:11124-30.
148 Bizon JL, Foster TC, Alexander GE, Glisky EL. 2012. Characterizing cognitive aging of working memory and executive function in animal models. Front Aging Neurosci 4:19. Bliss TV, Lomo T. 1973. Long-lasting potentiation of synaptic transmission in the dentate area of the anaesthetized rabbit following stimulation of the perforant path. J Physiol 232:331-56. Bloss EB, Janssen WG, Ohm DT, Yuk FJ, Wadsworth S, Saardi KM, McEwen BS, Morrison JH. 2011. Evidence for reduced experience-dependent dendritic spine plasticity in the aging prefrontal cortex. J Neurosci 31:7831-9. Boisvert MM, Erikson GA, Shokhirev MN, Allen NJ. 2018. The Aging Astrocyte Transcriptome from Multiple Regions of the Mouse Brain. Cell Rep 22:269-285. Bonder DE, McCarthy KD. 2014. Astrocytic Gq-GPCR-linked IP3R-dependent Ca2+ signaling does not mediate neurovascular coupling in mouse visual cortex in vivo. J Neurosci 34:13139-50. Bosch M, Hayashi Y. 2012. Structural plasticity of dendritic spines. Curr Opin Neurobiol 22:383-8. Bourin M, Hascoet M. 2003. The mouse light/dark box test. Eur J Pharmacol 463:55-65. Bourin M, Petit-Demouliere B, Dhonnchadha BN, Hascoet M. 2007. Animal models of anxiety in mice. Fundam Clin Pharmacol 21:567-74. Bourne J, Harris KM. 2007. Do thin spines learn to be mushroom spines that remember? Curr Opin Neurobiol 17:381-6. Bowser DN, Khakh BS. 2007. Vesicular ATP is the predominant cause of intercellular calcium waves in astrocytes. J Gen Physiol 129:485-91. Burke SN, Barnes CA. 2006. Neural plasticity in the ageing brain. Nat Rev Neurosci 7:30-40. Bushong EA, Martone ME, Jones YZ, Ellisman MH. 2002. Protoplasmic astrocytes in CA1 stratum radiatum occupy separate anatomical domains. J Neurosci 22:183-92. Cabezas R, Avila-Rodriguez M, Vega-Vela NE, Echeverria V, Gonzalez J, Hidalgo OA, Santos AB, Aliev G, Barreto GE. 2016. Growth Factors and Astrocytes Metabolism: Possible Roles for Platelet Derived Growth Factor. Med Chem 12:204-10. Cahoy JD, Emery B, Kaushal A, Foo LC, Zamanian JL, Christopherson KS, Xing Y, Lubischer JL, Krieg PA, Krupenko SA, Thompson WJ, Barres BA. 2008. A transcriptome database for astrocytes, neurons, and oligodendrocytes: a new resource for understanding brain development and function. J Neurosci 28:264-78. Cai J, Chen Y, Cai WH, Hurlock EC, Wu H, Kernie SG, Parada LF, Lu QR. 2007. A crucial role for Olig2 in white matter astrocyte development. Development 134:1887-99. Calnan DR, Brunet A. 2008. The FoxO code. Oncogene 27:2276-88. Cao X, Li LP, Wang Q, Wu Q, Hu HH, Zhang M, Fang YY, Zhang J, Li SJ, Xiong WC, Yan HC, Gao YB, Liu JH, Li XW, Sun LR, Zeng YN, Zhu XH, Gao TM. 2013. Astrocyte-derived ATP modulates depressive-like behaviors. Nat Med 19:773-7. Castagne V, Moser P, Roux S, Porsolt RD. 2011. Rodent models of depression: forced swim and tail suspension behavioral despair tests in rats and mice. Curr Protoc Neurosci Chapter 8:Unit 8 10A. Castelhano-Carlos MJ, Sousa N, Ohl F, Baumans V. 2010. Identification methods in newborn C57BL/6 mice: a developmental and behavioural evaluation. Lab Anim 44:88-103. Cenquizca LA, Swanson LW. 2007. Spatial organization of direct hippocampal field CA1 axonal projections to the rest of the cerebral cortex. Brain Res Rev 56:1-26. Cerqueira JJ, Pego JM, Taipa R, Bessa JM, Almeida OF, Sousa N. 2005. Morphological correlates of corticosteroid-induced changes in prefrontal cortex-dependent behaviors. J Neurosci 25:7792800. Chai H, Diaz-Castro B, Shigetomi E, Monte E, Octeau JC, Yu X, Cohn W, Rajendran PS, Vondriska TM, Whitelegge JP, Coppola G, Khakh BS. 2017. Neural Circuit-Specialized Astrocytes: Transcriptomic, Proteomic, Morphological, and Functional Evidence. Neuron 95:531-549 e9.
149 Charles AC, Merrill JE, Dirksen ER, Sanderson MJ. 1991. Intercellular signaling in glial cells: calcium waves and oscillations in response to mechanical stimulation and glutamate. Neuron 6:983-92. Chen J, Tan Z, Zeng L, Zhang X, He Y, Gao W, Wu X, Li Y, Bu B, Wang W, Duan S. 2013. Heterosynaptic long-term depression mediated by ATP released from astrocytes. Glia 61:178-91. Chen N, Sugihara H, Sharma J, Perea G, Petravicz J, Le C, Sur M. 2012. Nucleus basalis-enabled stimulus-specific plasticity in the visual cortex is mediated by astrocytes. Proc Natl Acad Sci U S A 109:E2832-41. Ciocchi S, Passecker J, Malagon-Vina H, Mikus N, Klausberger T. 2015. Brain computation. Selective information routing by ventral hippocampal CA1 projection neurons. Science 348:560-3. Clapcote SJ, Lazar NL, Bechard AR, Wood GA, Roder JC. 2005. NIH Swiss and Black Swiss mice have retinal degeneration and performance deficits in cognitive tests. Comp Med 55:310-6. Clarke LE, Liddelow SA, Chakraborty C, Munch AE, Heiman M, Barres BA. 2018. Normal aging induces A1-like astrocyte reactivity. Proc Natl Acad Sci U S A 115:E1896-E1905. Copeland CS, Wall TM, Sims RE, Neale SA, Nisenbaum E, Parri HR, Salt TE. 2017. Astrocytes modulate thalamic sensory processing via mGlu2 receptor activation. Neuropharmacology 121:100-110. Cornell-Bell AH, Finkbeiner SM, Cooper MS, Smith SJ. 1990. Glutamate induces calcium waves in cultured astrocytes: long-range glial signaling. Science 247:470-3. Cotrina ML, Nedergaard M. 2002. Astrocytes in the aging brain. J Neurosci Res 67:1-10. Cotter D, Mackay D, Chana G, Beasley C, Landau S, Everall IP. 2002. Reduced neuronal size and glial cell density in area 9 of the dorsolateral prefrontal cortex in subjects with major depressive disorder. Cereb Cortex 12:386-94. Covelo A, Araque A. 2016. Lateral regulation of synaptic transmission by astrocytes. Neuroscience 323:62-6. Cryan JF, Holmes A. 2005. The ascent of mouse: advances in modelling human depression and anxiety. Nat Rev Drug Discov 4:775-90. Dallerac G, Rouach N. 2016. Astrocytes as new targets to improve cognitive functions. Prog Neurobiol 144:48-67. Das MM, Svendsen CN. 2015. Astrocytes show reduced support of motor neurons with aging that is accelerated in a rodent model of ALS. Neurobiol Aging 36:1130-9. de Kloet ER, Joels M, Holsboer F. 2005. Stress and the brain: from adaptation to disease. Nat Rev Neurosci 6:463-75. De Pitta M, Brunel N, Volterra A. 2016. Astrocytes: Orchestrating synaptic plasticity? Neuroscience 323:43-61. Derouiche A, Frotscher M. 2001. Peripheral astrocyte processes: monitoring by selective immunostaining for the actin-binding ERM proteins. Glia 36:330-41. Di Castro MA, Chuquet J, Liaudet N, Bhaukaurally K, Santello M, Bouvier D, Tiret P, Volterra A. 2011. Local Ca2+ detection and modulation of synaptic release by astrocytes. Nat Neurosci 14:127684. Dias-Ferreira E, Sousa JC, Melo I, Morgado P, Mesquita AR, Cerqueira JJ, Costa RM, Sousa N. 2009. Chronic stress causes frontostriatal reorganization and affects decision-making. Science 325:621-5. Dickstein DL, Kabaso D, Rocher AB, Luebke JI, Wearne SL, Hof PR. 2007. Changes in the structural complexity of the aged brain. Aging Cell 6:275-84. Dimou L, Gotz M. 2014. Glial cells as progenitors and stem cells: new roles in the healthy and diseased brain. Physiol Rev 94:709-37. Djukic B, Casper KB, Philpot BD, Chin LS, McCarthy KD. 2007. Conditional knock-out of Kir4.1 leads to glial membrane depolarization, inhibition of potassium and glutamate uptake, and enhanced short-term synaptic potentiation. J Neurosci 27:11354-65.
150 Dolgin E. 2013. Old mice require new experimental tricks to study aging process. Nat Med 19:518-9. Dong QP, He JQ, Chai Z. 2013a. Astrocytic Ca(2+) waves mediate activation of extrasynaptic NMDA receptors in hippocampal neurons to aggravate brain damage during ischemia. Neurobiol Dis 58:68-75. Dong Z, Bai Y, Wu X, Li H, Gong B, Howland JG, Huang Y, He W, Li T, Wang YT. 2013b. Hippocampal long-term depression mediates spatial reversal learning in the Morris water maze. Neuropharmacology 64:65-73. Durkee CA, Araque A. 2019. Diversity and Specificity of Astrocyte-neuron Communication. Neuroscience 396:73-78. Edlund T, Jessell TM. 1999. Progression from extrinsic to intrinsic signaling in cell fate specification: a view from the nervous system. Cell 96:211-24. Eijkelenboom A, Burgering BM. 2013. FOXOs: signalling integrators for homeostasis maintenance. Nat Rev Mol Cell Biol 14:83-97. Emsley JG, Macklis JD. 2006. Astroglial heterogeneity closely reflects the neuronal-defined anatomy of the adult murine CNS. Neuron Glia Biol 2:175-86. Engle JR, Barnes CA. 2012. Characterizing cognitive aging of associative memory in animal models. Front Aging Neurosci 4:10. Fanselow MS, Dong HW. 2010. Are the dorsal and ventral hippocampus functionally distinct structures? Neuron 65:7-19. Farmer WT, Murai K. 2017. Resolving Astrocyte Heterogeneity in the CNS. Front Cell Neurosci 11:300. Feriod CN, Nguyen L, Jurczak MJ, Kruglov EA, Nathanson MH, Shulman GI, Bennett AM, Ehrlich BE. 2014. Inositol 1,4,5-trisphosphate receptor type II (InsP3R-II) is reduced in obese mice, but metabolic homeostasis is preserved in mice lacking InsP3R-II. Am J Physiol Endocrinol Metab 307:E1057-64. Fiacco TA, McCarthy KD. 2018. Multiple Lines of Evidence Indicate That Gliotransmission Does Not Occur under Physiological Conditions. J Neurosci 38:3-13. Fields RD, Araque A, Johansen-Berg H, Lim SS, Lynch G, Nave KA, Nedergaard M, Perez R, Sejnowski T, Wake H. 2014. Glial biology in learning and cognition. Neuroscientist 20:426-31. Foley J, Blutstein T, Lee S, Erneux C, Halassa MM, Haydon P. 2017. Astrocytic IP3/Ca2+ Signaling Modulates Theta Rhythm and REM Sleep. Front Neural Circuits 11:3. Fox E. 2008. Emotion Science: Cognitive and Neuroscientific Approaches to Understanding Human Emotions: Palgrave Macmillan. Fox WM. 1965. Reflex-ontogeny and behavioural development of the mouse. Anim Behav 13:234-41. Freeman SH, Kandel R, Cruz L, Rozkalne A, Newell K, Frosch MP, Hedley-Whyte ET, Locascio JJ, Lipsitz LA, Hyman BT. 2008. Preservation of neuronal number despite age-related cortical brain atrophy in elderly subjects without Alzheimer disease. J Neuropathol Exp Neurol 67:1205-12. Futatsugi A, Kato K, Ogura H, Li ST, Nagata E, Kuwajima G, Tanaka K, Itohara S, Mikoshiba K. 1999. Facilitation of NMDAR-independent LTP and spatial learning in mutant mice lacking ryanodine receptor type 3. Neuron 24:701-13. Futatsugi A, Nakamura T, Yamada MK, Ebisui E, Nakamura K, Uchida K, Kitaguchi T, Takahashi-Iwanaga H, Noda T, Aruga J, Mikoshiba K. 2005. IP3 receptor types 2 and 3 mediate exocrine secretion underlying energy metabolism. Science 309:2232-4. Galeotti N, Vivoli E, Norcini M, Bartolini A, Ghelardini C. 2008. An antidepressant behaviour in mice carrying a gene-specific InsP3R1, InsP3R2 and InsP3R3 protein knockdown. Neuropharmacology 55:1156-64. Ge Y, Dong Z, Bagot RC, Howland JG, Phillips AG, Wong TP, Wang YT. 2010. Hippocampal long-term depression is required for the consolidation of spatial memory. Proc Natl Acad Sci U S A 107:16697-702.
151 Giaume C, Koulakoff A, Roux L, Holcman D, Rouach N. 2010. Astroglial networks: a step further in neuroglial and gliovascular interactions. Nat Rev Neurosci 11:87-99. Gibbs ME, Hutchinson D, Hertz L. 2008. Astrocytic involvement in learning and memory consolidation. Neurosci Biobehav Rev 32:927-44. Gittins RA, Harrison PJ. 2011. A morphometric study of glia and neurons in the anterior cingulate cortex in mood disorder. J Affect Disord 133:328-32. Gomez-Gonzalo M, Martin-Fernandez M, Martinez-Murillo R, Mederos S, Hernandez-Vivanco A, Jamison S, Fernandez AP, Serrano J, Calero P, Futch HS, Corpas R, Sanfeliu C, Perea G, Araque A. 2017. Neuron-astrocyte signaling is preserved in the aging brain. Glia 65:569-580. Gomez-Gonzalo M, Navarrete M, Perea G, Covelo A, Martin-Fernandez M, Shigemoto R, Lujan R, Araque A. 2015. Endocannabinoids Induce Lateral Long-Term Potentiation of Transmitter Release by Stimulation of Gliotransmission. Cereb Cortex 25:3699-712. Gosselin RD, Gibney S, O'Malley D, Dinan TG, Cryan JF. 2009. Region specific decrease in glial fibrillary acidic protein immunoreactivity in the brain of a rat model of depression. Neuroscience 159:91525. Gosselin T, Le Guisquet AM, Brizard B, Hommet C, Minier F, Belzung C. 2017. Fluoxetine induces paradoxical effects in C57BL6/J mice: comparison with BALB/c mice. Behav Pharmacol 28:466476. Gourine AV, Kasymov V, Marina N, Tang F, Figueiredo MF, Lane S, Teschemacher AG, Spyer KM, Deisseroth K, Kasparov S. 2010. Astrocytes control breathing through pH-dependent release of ATP. Science 329:571-5. Grady CL. 2008. Cognitive neuroscience of aging. Ann N Y Acad Sci 1124:127-44. Granes F, Urena JM, Rocamora N, Vilaro S. 2000. Ezrin links syndecan-2 to the cytoskeleton. J Cell Sci 113 ( Pt 7):1267-76. Graves AR, Moore SJ, Bloss EB, Mensh BD, Kath WL, Spruston N. 2012. Hippocampal pyramidal neurons comprise two distinct cell types that are countermodulated by metabotropic receptors. Neuron 76:776-89. Graziano A, Petrosini L, Bartoletti A. 2003. Automatic recognition of explorative strategies in the Morris water maze. J Neurosci Methods 130:33-44. Griebel G, Simiand J, Serradeil-Le Gal C, Wagnon J, Pascal M, Scatton B, Maffrand JP, Soubrie P. 2002. Anxiolyticand antidepressant-like effects of the non-peptide vasopressin V1b receptor antagonist, SSR149415, suggest an innovative approach for the treatment of stress-related disorders. Proc Natl Acad Sci U S A 99:6370-5. Gu Y, Arruda-Carvalho M, Wang J, Janoschka SR, Josselyn SA, Frankland PW, Ge S. 2012. Optical controlling reveals time-dependent roles for adult-born dentate granule cells. Nat Neurosci 15:1700-6. Guerra-Gomes S, Sousa N, Pinto L, Oliveira JF. 2017. Functional Roles of Astrocyte Calcium Elevations: From Synapses to Behavior. Front Cell Neurosci 11:427. Guerra-Gomes S, Viana JF, Nascimento DSM, Correia JS, Sardinha VM, Caetano I, Sousa N, Pinto L, Oliveira JF. 2018. The Role of Astrocytic Calcium Signaling in the Aged Prefrontal Cortex. Frontiers in Cellular Neuroscience 12. Gutierrez Y, Garcia-Marques J, Liu X, Fortes-Marco L, Sanchez-Gonzalez R, Giaume C, Lopez-Mascaraque L. 2019. Sibling astrocytes share preferential coupling via gap junctions. Glia 67:1852-1858. Habbas S, Santello M, Becker D, Stubbe H, Zappia G, Liaudet N, Klaus FR, Kollias G, Fontana A, Pryce CR, Suter T, Volterra A. 2015. Neuroinflammatory TNFalpha Impairs Memory via Astrocyte Signaling. Cell 163:1730-41.