Project Topics
Get Complete Project Material File(s) Now! »
Novels methods for investigating the functional connec-tivity
The major aim of systems neuroscience is to understand how the brain processes information, encodes perception and generates behavior. These operations are determined by the structure of the neural circuits, the physiological and anatomical features of their local and external connections as well as the integrative properties of neurons that compose them. Critical steps toward the comprehension of neuronal circuits are the identification of the inputs of the cir-cuit’s diverse neuron types, and an understanding of their interaction within the circuit. Sev-eral neuronal tracing techniques exist to chart anatomical connections within the nervous sys-tem. Anterograde and retrograde neuronal tracers have allowed us to gain information about connectivity between different cortical and subcortical areas. However, when used alone, they appear insufficient to reveal connectivity at finer levels of complexity. In recent years, a toolbox of genetically encoded molecules has emerged and has enabled us to target specific cell types and control their electrical activity in a temporally precise fashion by pulses of light. Modern tracing strategies combined with innovative optical methods work in concert to reveal multisynaptic pathways and allow to identify connections to and from particular cell types. In this section, I will briefly describe neuronal tracers such as Retrobeads, Cholera toxin B as well as the optical tools that I have used during my PhD.
Neuroanatomical tracers
Neuronal tracing techniques are based on the principle of axonal transport. With regard to the direction of transport, the notion of anterograde and retrograde transportation of tracer materi-al can be distinguished. For retrograde transport, the tracer material enters from the cell axons (usually by endocytosis) and is transported back to the cell body. For anterograde transport, the uptake mechanisms involve the cell soma and/or its dendrites, and the tracer is carried along the axonal microtubular system to the cell’s synaptic terminals.
Retrobeads
Fluorescently tagged microspheres or beads, first introduced by Katz et al., (1984), are 20-200nm large polystyrene (Molecular Probes) or latex (Lumafluor) microspheres labeled with dyes of different colors. Those from Lumafluor are red (rhodamine-labeled) or green (fluores-cein-labelled) (Katz et al., 1984; Katz and Iarovici, 1990). They are suitable to trace both local and long-range connections, depending on the injected volume of tracer solution. They diffuse very little into the surrounding brain tissue and produce sharply defined injection sites. Rhodamine and fluorescein are fluorescent proteins that are soluble in water, as well as etha-nol and methanol. The microbeads are taken up by axon terminals within the injection site and are transported retrogradely into the soma. The uptake mechanism is not completely under-stood (possibly a latex-triggered endocytosis); however, size and surface properties seem to play an important role. Retrograde transport is fast since the labelling increases until 48h after injection and is then persistent for several weeks. In retrogradely labeled pyramidal neurons, beads fill the cell soma and the proximal trunk of the apical dendrites. Latex microspheres are not cytotoxic for the animal, and no phototoxicity for labeled cells after illumination has been described. Latex microspheres can also be combined with immunohistochemistry, intracellu-lar injections, immunofluorescence, Golgi silver impregnation or in situ hybridization. Under illumination, the fluorescence is stable and latex beads are directly visible without the need of further staining techniques, which makes them suitable for in vitro, in vivo and cell culture experiments. The possibility to record from labeled neurons or identify specific afferents makes the beads a marker of choice for functional connectivity studies. Severed fibers of pas-sage, but not intact ones, seem to show some uptake of marker material. Once sectioned and mounted, beads fade over time because of their sensitivity to glycerol unless antifading spe-cific mounting media are used. Beads in retrogradely labelled cells appear as granular dots in the cytoplasm. The rhodamine and fluorescein components are best excited with a light source of 540nm and 490nm wavelength respectively, while the maximum emission lies at 590nm and 520nm for red and green labelling respectively (Katz et Iarovici, 1990). Due to their over-lapping wavelengths, cells brightly labelled with red beads can, however, interfere with the identification of cells that are more weakly labelled with green beads (Köbbert et al., 2000).
Cholera toxin B
Cholera toxin is comprised of two subunits, A and B, forming a hexameric complex. The A subunit consists of a single copy and is considered as a ADP-ribosyltransferase enzyme, which disrupts the G protein signaling leading eventually to dehydration of the cell (Finkel-stein and Dorner, 1985). The nontoxic B subunit arranged as a pentameric ring is important to the protein complex as it allows the protein to bind to cellular surfaces via the pentasaccharide chain of ganglioside GM1 membrane receptor facilitating its internalization. Cholera toxin B (CTB) is an axonally transported marker and has proven to be a powerful tool for retrograde labeling of neurons. It has been recently used as a marker of lipid rafts (membrane microdo-mains enriched in cholesterol and sphingolipids) involved in cell signaling and protein traf-ficking (Janes et al., 1999). Historically, CTB has been limited to single-labeling techniques using bright-field horseradish peroxidase and DAB staining (Luppi et al., 1990). Recently, CTB conjugated with fluorescein- or rhodamine fluorochromes has been used to trace the connections of multiple brain areas. Fluorescently conjugated-CTBs are made from the pure original recombinant cholera toxin subunit B and completely free of the toxic A subunit (Conte et al., 2009a). Even at low concentrations (0.5%), dye-labeled CTB is very effective and sensitive and only a slight increase in background labeling was observed at higher con-centrations (1%) (Conte et al., 2009b). Labeling is photostable and resists photobleaching for several months after coverslipping. After axonal transport, labeling is preferentially localized to soma and proximal dendrites. Because CTB remains in vesicles, labeled somata appear granular. Labelling is optimal after 2-4 days of survival even though it varies from 1 to 10 days depending on the age of the animal (Angelucci et al., 1996). CTB may be taken up by non-terminal passing fibers (Chen and Jones, 1995), which is a common problem for most neuronal tracers. Mostly considered as a retrograde tracer, some anterograde transport has been described (Angelucci et al., 1996).
Optogenetics
Electrical stimulation has been widely used with great efficacy to both control and probe the function of discrete brain regions. However, it is not possible to target genetically specified cell types, a disadvantage that can be overcome with genetically encoded molecules. A recent innovation in experimental neuroscience has been the development of light-activated channels or pumps, derived from microbial photosynthetic systems, to modulate neural activity. The expression of these photosensitive molecules is driven through genetic manipulation of the target cells, which is why these tools are referred to as optogenetics. These optogenetics tools allow for activation or silencing of neurons with high temporal precision and specificity. One frequently used approach to express light sensitive molecules in neurons is the use of lentivi-rus or adeno-associated virus (AAV) for viral transduction in vivo. During my PhD, I used AAV virus vectors for in vivo delivery of plasmids containing ChR2 and a fluorescent report-er. Delivered through stereotactic injection in the brain, this method allows to target a spatial-ly restricted brain area. Cell type specificity depends on the serotype and promoter used. This approach can lead to high level of protein expression within short periods of time.
Adeno-associated virus
AAV is a 20-30 nm diameter non-enveloped (no phospholipid coat), single-stranded, small DNA virus belonging to the Parvovirus family. It is a non-pathogenic virus with a genome of approximately 4.7kb flanked by two 145 base inverted terminal repeats (ITRs) on the 5’ and 3’ ends. The AAV genome consists of two open reading frames (ORFs) that encode four replication proteins (Rep), three capsid proteins (Cap/VP) and other assembly proteins (AAPs) (Murlidharan et al., 2014). ITRs are the only cis-acting elements necessary for successful encapsidation and replication of the virus. AAV provides a source for recombinant gene vectors derived from AAV (or rAAV) that carry no protein coding sequences. Wild type AAV virus requires co-infection by adenovirus or herpes simplex virus for efficient replication in their host, and will, in the absence of helper virus, stably integrate into the host cell genome. AAV vector genomes remain primarily episomal in target cells and have a low (if any) frequency of integration (https://www.med.upenn.edu/ gtp/vectorcore/BiosafetyInformation.shtml). AAV demonstrates relatively low immunogenicity, and can be produced in large quantities with high purity and thus, is useful for long-term and high level of expression in the nervous system. Compared to lentiviral vectors, AAVs lead to greater spatial spread due to their small particle diameter (100nm compared to 20-30nm for AAVs). AAVs enter the cell through interaction between the capsid and the AAV membrane receptors (Pillay et al., 2016), preferentially the cell surface glycans (Asokan et al., 2012), which leads to vector internalization (Fig. 25; Russel and Kay, 1999). The vector DNA is then released in the cytosol and translocated to the nucleus (Russel and Kay, 1999). The different serotypes have different sequences in their ITRs and in their capsid proteins. The ITR serotype is the number listed first in the name of an AAV vector, while the capsid type is the second number. AAV serotypes noted from AAV2/1 to AAV2/9 exhibit a range of properties related to antigenicity, in vivo tropism, and receptor interactions based on their different capsid structures (Burger et al., 2005). In addition, the speed of onset of viral gene expression for maximal number of cells can vary up to several weeks or months among serotypes (Aschauer et al., 2013). Opsins are found in an extensive variety of organisms ranging from archaebacteria to mam-mals. These proteins are composed of a seven-transmembrane-helix-domain (7TM) covalent-ly linked retinal, the aldehyde of vitamin A. Opsins are separated into two groups, depending on their primary sequence, type I (microbial) and type 2 (animal). Microbial-type opsins are found in archaea, eubacteria, fungi and algae, whereas animal-type 2 opsins are encountered in the mammalian retina photoreceptor cells. Optogenetics takes advantage of this dissocia-tion by expressing de novo microbial opsins, for example in mammalian cells, conferring them new light sensitivity. Two classes of microbial opsins have been described: the light-driven ion pumps, including the chloride pumps halorhodopsins (HR) and the proton pumps bacteriorhodopsins (BR), and the light-driven ion channels channelrhodopsins (ChR) (Fig. 26). The first microbial opsin applied both in vitro and in vivo to excite neurons was chan-nelrhodopsin-2 (ChR2), a member of the channelrhodopsin (ChR) family. Channelrhodopsins are light-gated nonspecific cation channels, conducting Na+, K+ and Ca2+ ions, derived from the green algae C. Reinhardtii (Nagel et al., 2003). Channelrhodopsins possess a molecule of all-trans retinal (ATR) bound at the core as a photosensor. Upon illumination with maximum excitation at 470 nm (blue light), ATR isomerizes and triggers a conformational change to open the channel pore, thus depolarizing expressing neurons and inducing single action poten-tials (Boyden et al., 2005). The main limitation of ChR2 is the high level of desensitization that decays by 80% from the peak to a steady-state response (Nagel et al., 2003). At high ex-pression levels, ChR2 can form intracellular aggregates instead of trafficking to the mem-brane, extra spikes can occur (two spikes following one light pulse) and cells can enter “depo-larization block” (repetitive stimulation does not allow repolarization of the membrane). A variant of ChR2 with a mutation at position H134R leads to a modest reduction in desensitiza-tion, but also to a slight increase in light sensitivity and slower channel closing, favoring de-polarization block (Nagel et al., 2005; Lin et al., 2009). The E123T mutation in ChR2 (ChETA) creates faster kinetics, reduces extra-spikes and allows for temporary sustained spike trains up to 200Hz (Gunaydin et al., 2010). ChR2 application was quickly followed by the development of the inhibitory halorhodopsin (Chow et al., 2010) that could hyperpolarize and therefore silence expressing neurons upon illumination with 580 nm yellow light via an inward chloride flux with NpHR derived from the archaeon Natronomonas pharaonic. Multi-channel photoinhibition and photostimulation can drive high-fidelity sequences of hyperpolar-izations and depolarizations in neurons simultaneously expressing yellow light-driven HR and blue light-driven ChR2, allowing for the first time manipulations of neural synchrony without perturbation of other parameters such as spiking rates (Han and Boyden, 2007). The light-activated Halorubrum sodomense archaerhodopsin (Arch) is a light-driven proton pump. Once expressed in neurons and activated with yellow or green light, Arch pumps positive charge H+ out of the cells, hyperpolarizing them. It possesses strong photocurrents and can mediate complete silencing of neuronal activity in awake behaving mice (Chow et al., 2012). For ease of identification of transfected cells, optogenetic molecules can be tagged by fluores-cent marker proteins such as eYFP, mCherry or tdTomato, on its C-terminal..
Aims of the thesis
The presubiculum is a transitional cortical area that is part of the parahippocampal region. As presented in the introduction, despite its functional importance in spatial coding, many aspects of the structural and functional organization of the presubiculum have remained poorly understood. Specifically, inhibitory components and their functional connectivity within the microcircuit are still unexplored. My supervisor, Desdemona Fricker, therefore proposed me a thesis project whose aim was to elucidate the information processing at the level of the inhibitory microcircuit in mouse presubiculum. The presubiculum contains head direction cells, which fire as a function of the animal’s directional heading. The anterior thalamic nuclei, together with the visual and retrosplenial cortex are major inputs to the presubiculum and greatly contribute to the generation and the refinement of the head direction signal in this area. The role of the presubiculum is to distribute a visual landmark control to subcortical areas and it is considered a major drive to the downstream entorhinal cortex where it contributes to spatially tuned firing. It is likely that information in the presubiculum is actively integrated and refined. In this framework, the aims of my thesis were 1) to determine the cellular components and 2) the local and afferent connectivity of the inhibitory microcircuit, 3) to contribute experimental data on connectivity and synaptic properties of interneurons for computational models and to gain insight in the function of GABAergic interneurons in HD signaling. To achieve this, I took advantage of transgenic mouse lines that specifically label PV and SOM interneurons, the PVCre::dtTomato, SSTCre::dtTomato and X98 GFP to study the main inhibitory elements of the microcircuit. Using dual or single whole-cell patch-clamp recordings in combination with optogenetic tools that specifically activated afferent inputs and retrograde tracers that label presubicular efferent projections, I wished to examine the anatomy, electrophysiology and the input-output connectivity of inhibitory presubicular neurons.
Slices
During my PhD, I was always striving to improve my slicing procedures. In order to perform electrophysiological recordings, the quality of slices was a crucial point and depended on the age of the animal. From the preparation of solutions to the dissection and slicing, every step has to be controlled and performed meticulously. Many recipes and advices are found in www.BrainSliceMethods.com. Animals were anesthetized with ketamine hydrochloride (Imalgene®) and xylazine (Rompun®) (100 and 15 mg.kg−1, respectively). The mixture is prepared as followed: I first added 1mL of Imalgene and 0.5 mL of Rompun, then I added 8.5 mL of NaCl to get a final volume (10mL) of 1/8 diluted solution. The injected volume depended on the weight of the animal (100µL of mixture per 10g). 100 µL of Choay heparine was injected intraperitoneally on the opposite side after anesthesia. When the mouse was deeply anesthetized (no response to nociceptive stimulation), another 100 µL of Choay heparine was injected in the left ventricle before perfusion and the descending aorta was clamped. Animals were then perfused through the heart by gravity flow with a sucrose-based solution with 125 NaCl, 25 sucrose, 2.5 KCl, 25 NaHCO3, 1.25 NaH2PO4, 2.5 D-glucose, 0.1 CaCl2, 7 MgCl2 (in mM) cooled to 4◦C and equilibrated with 5% CO2 in O2. A perfusion needle was inserted into the left ventricle and after 5 seconds, the right atrium was cut with small scissors. A perfusion was judged to have worked when the animal lungs and rib cage bones turned white. The volume of perfusion varied from 20 to 50 mL. The speed was adjusted so that the solution came out of the perfusion needle by dripping. The stock solution could be kept at 4°C for two months. Cutting solution was prepared from 100 mL of 1L of 10X stock solution. Then, it was put in ice or in the freezer to decrease the temperature until 0-4°C. The day of the experiment, I started to oxygenate the cutting solution for at least 1 minute before adding 0.1 mM CaCl2 and 7 mM of MgCl2 in order to avoid their precipitation. CaCl2 and MgCl2 were always added in the final cutting solution but never in the stock solution. pH and osmolarity were then measured. Sucrose-containing solution was prepared from MilliQ water (Elix®, Millipore; resistance of 18.2µΩ). A good perfusion significantly improves the slice quality with animals older than 25 days. Animals were decapitated and horizontal 280–320µm thick brain sections were cut in the same solution using a vibratome (Leica VT1000S or Microm HM650V). To obtain horizontal slices, the cerebellum was chopped off, then the dorsal part of the brain was stuck on the slic-er platform using superglue. Cutting started from the ventral side, the parahippocampal area with the entorhinal cortex turned towards the blade. Some ice was put around the chamber to keep the cutting solution as cold as possible during slicing (usually between 0-4°C). Compared to coronal and parasagittal angles, horizontal slicing appeared more appropriate for preserving neuronal health and integrity of the presubicular tissue. Slices were stored for 20 min at 37°C in a holding immerged chamber filled with ACSF containing (in mM): 124 NaCl, 2.5 KCl, 26 NaHCO3, 1 NaH2PO4, 2 CaCl2, 2 MgCl2, and 11 D-glucose, bubbled with 5% CO2 in O2 (pH 7.3, 305–315 mOsm/L). Storing at physiological temperatures appeared to improve slicing quality for older animals. Then, they were stored for at least 45 min at room temperature in the same solution gently bubbled with 5% CO2 in O2. I noticed that warming and storing acute slices in an interface chamber (designed with the help of Marie Goutierre, a PhD student in Jean Christophe Poncer lab) significantly improved neuronal health and facilitated my recordings from retrobeads-labeled neurons. The stock solution was kept at 4°C for two months. All components (listed in table 2) except CaCl2 and MgCl2 were added in the final stock solution. Cutting solution was prepared from 100 mL of 1L 10X stock solution only the day of experiment. Before adding 2 mM of CaCl2 and 2 mM of MgCl2, I started to oxygenate the solution for at least 1 minute to avoid their precipitation. pH (7.3) and osmolarity (305–315 mOsm/L) were then measured. ACSF was prepared from MilliQ water (Elix®, Millipore; resistance of 18.2µΩ).
Patch-clamp recordings
Slices were then transferred to a recording chamber (volume 2–3 mL, temperature 33–35◦C) mounted on a BX51WI microscope (Olympus, France). Fluorescently labeled PV, SST or X98 interneurons were identified using LED illumination with appropriate emis-sion/excitation filters (OptoLED, blue (470nm) and yellow (580nm) LED, Cairn Research, Faversham, UK). Cf. figure 27 for additional information. Recordings were made with glass pipettes pulled using a Brown–Flaming electrode puller (Sutter Instruments) from borosilicate glass of external and internal diameter of 1.5 mm and 0.86, respectively (Harvard Apparatus, UK; reference: GC150F-10). The electrode resistance, when filled with the internal solution was 3–8 MΩ. Dual recordings were easier to perform with higher resistance pipettes. For electrophysiological recordings, we used a MultiClamp 700B amplifier, a Digidata 1440a, a ISO-S-1.5G microelectrode holder (G23 instruments), pCLAMP software (Molecu-ar Devices, Union City, CA, USA), Luigs and Neumann micromanipulators and a heated chamber. The motorized table was commanded by Morgentau software (custom written by Michael Bendels, Bendels et al., 2008), used to position a lattice of light stimulation sites across the presubiculum. During my PhD, I used one kind of internal solution. The potassium-gluconate solution contained (in mM) 135 K-gluconate, 1.2 KCl, 10 HEPES, 0.2 ethylene glycol tetraacetic acid (EGTA), 2 MgCl2, 4 MgATP, 0.4 Tris-GTP, and 10 Na2-phosphocreatine. The internal solution stock (25mL) was prepared from distilled water (GibcoTM, Life technol-ogies) in cold temperature (4°C) to keep ATP and GTP stable. The components were added in the same order as listed in table 3. pH (7.3) was then measured and adjusted with KOH if nec-essary. Osmolarity (290 mOsm/L) was measured thereafter and adjusted with distilled water (if too high) or K-gluconate (if too low). Aliquots of 1 mL were then stored at -20°C. Fresh aliquots were always used for each day of experiment. Recordings were usually very stable with this internal solution after performing a gigaseal followed by a whole cell patch. Low EGTA did not alter stability and allowed calcium de-pendent physiological modifications. In order to study the inhibitory/excitatory conductance, 2mM QX 314 bromide (Tocris) was added the day of the experiment to an aliquot of 1mL of internal solution (and agitated), thus allowing to depolarize membrane potential to +40mV and eliminating action current in voltage-clamp mode. In order to perform double recordings, the two pipettes were positioned at the top of the slice, just above the targeted neurons that were carefully chosen by eye. Then, I patched one cell after another in whole-cell configura-tion. The protocol for testing the connectivity is described in Nassar et al. (in prep).
Analysis of electrophysiological data
Details concerning the analysis of the electrophysiological data will not be fully described here, as they are specifically depicted in the methods of the different studies. Electrophysio-logical parameters were analyzed with the following softwares: AxoGraph X (http://www.axograph.com/) for manual analysis, Spikoscope (a Labview based software de-veloped by Ivan Cohen) and routines developed with Matlab (the Mathworks) by Jean Si-monnet for computerized analysis.
Unsupervised cluster analysis for neuronal classification
As previously reviewed in the introduction section, interneurons can be classified according to many parameters such as the molecular content, morphology, intrinsic electrophysiological properties, their post-synaptic targets, their recruitment by specific inputs or as well as their specific behavioral-related activity (Ascoli et al., 2008). Classification of neocortical interneurons is a crucial step in understanding cortical circuits as each subtype of interneuron likely has a different function. Classifications have been based on one or a combination of descriptors chosen arbitrarily, often using qualitative criteria (e.g. shapes of the axo-dendritic tree or somatic layer location). Therefore, interneuron classifica-tions can change depending on which descriptors are used by the experimenter. It appears difficult to know which set of descriptors are the most relevant to determine a neuronal class and, more generally, how many classes of interneurons actually exist. Unsupervised clustering analysis is becoming a widely used method to identify neuronal clas-ses without predefining an initial group. This method uses algorithms to seek objective classi-fication of neuronal populations. It classifies objects by attributing the same weight to each parameter, thus, facilitating the grouping with a standardized nomenclature. Each neuron is thus represented by one point in a multidimensional space (the number of dimension equals the number of parameters) and the closest neurons in this space are then grouped together. I collaborated with Jean Simonnet to perform unsupervised cluster analysis in order to classi-fy presubicular interneurons (Nassar et al., 2015). Interneurons were grouped based on simi-larities of 17 electrophysiological parameters obtained from 159 recorded neurons. Our clus-tering was based on Ward’s method (Ward Jr, 1963), as previously used to classify neuronal populations (Cauli et al., 2000). Cluster analysis was implemented using the statistics toolbox of MATLAB (The Mathwork). Centering and reducing all values allowed the standardization of all values. At each step, the two nearest points that represented neurons were associated together using the matrix of their Euclidean distances and then used for the consecutive steps. The mean within-cluster distance was calculated. This value typically decreased as the num-ber of clusters increased. The Thorndike procedure (Thorndike, 1953), where jumps in dis-tances within clusters indicate prominent differences between groups, was used to examine resulting clusters. Final clusters were defined from statistical and biological parameters.
Neuronal morphology
To examine the morphology, I always added 1-3 mg/mL of biocytin in the internal solution for the patch-clamp recordings either in the internal stock solution preparation or in 1 aliquot during the day of the experiment. Biocytin was dissolved by agitation and no change in the pH and osmolarity was observed after measurement. To get a sufficiently complete axo-dendritic neuronal arborization, recordings should not exceed more than 20 min in order to avoid background staining. In addition, depending on the amount of biocytin, the staining can be more or less efficient. Indeed, the higher the amount of biocytin, the better is the staining. Slices containing filled cells were fixed in 4% paraformaldehyde (PFA) in 0.1M sodium phosphate buffer (PBS) overnight at 4°C. Slices were then rinsed in PBS (3×3 min) and cryo-protected in 30% sucrose at 4°C overnight. I noticed that it is possible to store slices in 30% sucrose for several weeks without altering the staining. Before starting biocytin revelation, membranes were permeabilized by three cycles of freez-ing/thawing. To perform this, slices were put on a slide, in a small drop of sucrose. Then, the slide was disposed on dry ice until sucrose is frozen. The experimenter palm was then used as a warm surface until full thawing. The procedure was repeated three times. Slices were then washed with PBS (2×30, then 1×60 min at slow agitation and room temperature) to complete-ly remove the remaining sucrose. Slices were agitated in saturation buffer containing 2% milk powder (Carnation) and 1% Triton X-100 in PBS 0.1M for 3 h at room temperature. The Tri-ton helps to permeabilize the membranes for the subsequent incubation. High concentration of Triton does not alter the staining and improves membrane permeabilization: slices are thick enough not to be degraded. Milk proteins are used to saturate the non-specific fixation sites, thus limiting the background staining. To reveal the biocytin staining, sections were gently agitated with Streptavidin–Cy3 or Cy5 conjugate (diluted at 1:500, Invitrogen, Eugene, OR, USA) and DAPI (1/1000) in the saturation buffer overnight at 4°C. DAPI is used as a nuclear marker to reveal tissue structure and define boundaries as well as layers of the presubiculum.
Table of contents :
1- The Hippocampal-Parahippocampal system: Anatomy, Connectivity and Function
1.1 Gross anatomy of the hippocampal-parahippocampal region
1.2 Hippocampal-Entorhinal connectivity
1.3 Spatial navigation and its neuronal schemes
2- The Presubiculum: Anatomy, Function and Connectivity
2.1 Anatomy of the presubicular cortex
2.1.1 Anatomical delineation
2.1.2 Laminar organization
2.1.3 Modular structures
2.2 Presubiculum and spatial orientation
2.2.1 Sense of orientation
2.2.2 Properties of presubicular head direction neurons
2.2.3 Head Direction Circuit
2.2.4 Subcortical source of head direction signal
2.2.5 Functional importance of presubiculum in spatial coding
2.3 Neuronal components of the presubiculum
2.3.1 Intrinsic excitability of presubicular neurons
2.3.2 Input and output regions of the presubicular microcircuit
3- GABAergic neurons of a cortical network
3.1 GABAergic neuron types
3.1.1 Morphological properties and postsynaptic targets
3.1.2 Molecular markers and gene expression
3.1.3 Physiology: firing patterns and intrinsic properties
3.2 Other characteristics of interneuron populations
3.2.1. Interneuron input connectivity
3.2.2. Synaptic excitation of interneurons
3.2.3. Interneuron outputs
3.3. Toward a classification of interneuron diversity
4-How inhibition shapes cortical information processing: functional importance of PV and SOM interneurons
4.1 Building blocks of the inhibitory circuit
4.2 The hippocampal-entorhinal circuit
4.2.1 Behavioral states and oscillations
4.2.2 Long-range GABAergic neurons
4.2.3 Importance of interneurons in grid cell activity
4.3 Visual processing
4.4 Somato-sensation
4.5 VIP interneurons – Influence of disinhibition
5-Novels methods for investigating the functional connectivity
5.1 Neuroanatomical tracers
5.1.1 Retrobeads
5.1.2 Cholera toxin B
5.2 Optogenetics
5.2.1 Adeno-associated virus
5.2.2 Microbial opsins
6-Aims of the thesis
