Catégorie d'événements: PhD defense

PhD defense in french

 Zoom link coming soon

Team : CAB

Supervisors : Pierre Moretto (CRCA-CBI) and Nicolas Turpin (IRISSE)

Committee members :

• BARBIER Frank (Rapporteur), Université Polytechnique Hauts-de-France, Laboratoire d’Automatique, de Mécanique et d’Informatique industrielles et Humaines (LAMIH UMR-CNRS 8201)
• CHEZE Laurence (Rapporteure), Université Gustave Eiffel, Laboratoire de biomécanique et mécanique des chocs (LBMC UMR_T9406)
• MARIN Frédéric (Examinateur), Université Technologie de Compiègne, Biomécanique et Bioingénierie (BMBI – UMR CNRS 7338)
• CHAVET Pascale (Examinatrice), Université Aix-Marseille, Institut des Sciences du Mouvement (ISM – UMR 7287)
• MORETTO Pierre (Directeur de thèse), Université Paul Sabatier, Centre de Recherches sur la Cognition Animale (CRCA – UMR 5169)
• TURPIN Nicolas (Codirecteur de thèse), Université de la Réunion, Ingénierie de la Santé, du Sport et de l’Environnement (IRISSE – UR 4075)

Abstract :

Collective load transport is a common task that we perform in our daily and professional lives. It involves the collaboration of two or more people to lift and move an object. Until now, the mechanisms underlying this movement and the collaboration have not been widely discussed in the literature. This thesis is part of the ANR CoBot project (Projet-ANR-18-CE10-0003) which aims to make a humanoid robot collaborative during load carrying. The objective of this work is to study the human-human interaction during a load transport, more particularly during a table or stretcher transport, and to extract biomechanical parameters that account for the control of this task. Three main questions motivate this project: (1) Does the participants’ individual and collective efficiency is altered with the task complexity? (2) Are locomotor involvement and strategies affected by backward and forward walking? and (3) Is the energy transfer occurring in the upper limbs efficient during load transport?

To answer these questions, the analysis carried out was based on techniques and methods already described in the literature and focused more specifically on: (i) an analysis of the locomotor pattern, (ii) an analysis of joint efforts and (iii) an analysis of energy transfer in the upper limbs. Our results show an alteration in the pendulum-like-behaviour of the participants’ centre of mass when adding a precision task to load transport. We also identified backward walking as the major factor impacting the locomotor patterns and efficiency in stretcher transport. The results show a distinct involvement and role for each participant. Depending on their placement and perceived environmental feedback, one participant will guide the movement, while the other will lift the load and follow the movement. These results are supported by the third study which shows that the participant who guides the movement generates and dissipates the energy required for handling the load, while the ‘follower’ is neutral. These results provide further insight into the interactions required during load transport and offer a wide range of potential applications. Indeed, the results can be used to improve control in cobotics, securing human-machine interactions, both during interactions with cobots and during tasks assisted by exoskeletons. Finally, our results should help to specify new ergonomic recommendations.

 

PhD defense in english

Zoom link coming soon

Team : IVEP

Supervisor : Jacques Gautrais (CRCA-CBI)

Committee members :

• Dr. Simon Benhamou – CEFE, Montpellier – Reviewer
• Dr. Carmen Bessa-Gomes – AgroParisTech, Paris – Reviewer
• Pr. Richard Fournier – LaPlace, Toulouse – Examiner
• Pr. Frederic Bartumeus – ICREA, Spain – Examiner
• Dr. Jacques Gautrais – Research Center on Animal Cognition, Toulouse – Thesis supervisor
• Dr. Alfonso Pérez-Escudero – Research Center on Animal Cognition, Toulouse – Guest
• Dr. Mathieu Lihoreau – Research Center on Animal Cognition, Toulouse – Guest

Abstract :

Understanding how pollinators move across space is key to understanding plant mating patterns. Bees are usually assumed to search for flowers randomly or using simple movement rules so that the probability of discovering a flower depends primarily on its distance to the nest. However, experimental work shows this is not always the case.

Until now, no one has successfully enunciated a realistic model of bee movement that considers the fact that they are Central Place Foragers and thus they start and end all their movements in the same place: the nest.

To further our knowledge of the exploratory movement of central place foraging bees, I propose a model of central place foraging that produces realistic bee trajectories by accounting for the autocorrelation of the bee’s angular speed, the attraction to the nest (homing), and Gaussian noise.

The four parameters of this model have been tuned based on experimental trajectories collected on bumblebees (Bombus terrestris) in the field. The model not only has the potential to describe the movement patterns of bees but also those of other central place forager animals.

The proposed model paves the way to compute theoretical predictions about pollination in the field. Here, I explored the statistics of flower discovery, depending on flower patch sizes and densities.

Simulations of bumblebee trajectories highlight two effects that were previously overlooked: a masking effect that reduces the detection of flowers close to another and a scale effect that modulates this first effect as a function of the distance between flowers.

At the plant level, flowers distant from the nest were more often discovered by bees in low-density environments. At the colony level, foragers found the most flowers when they were small and at medium densities. These results suggest that pollination would be optimized in a range of intermediate flower densities: when the density is too low, few flowers are discovered; when the density is too high, flowers distant from the nest become masked by closer ones (due to the masking effect).

These results indicate that the processes of search and discovery of resources are potentially more complex than usually assumed, and question the importance of resource distribution and abundance on bee foraging success and plant pollination.

© Crédit photo: Romain Hacquet – @aker.romain

 

PhD defense in french

Zoom link : https://univ-tlse3-fr.zoom.us/j/99150206269

Team : REMEMBER

Supervisor : Lionel MOULEDOUS (CRCA-CBI)

Committee members :

• Dr. Lionel Moulédous – CRCA
• Pr. Bruno Guiard – CRCA
• Dr. Catherine Mollereau-Manaute – CRCA
• Dr. Domonique Massotte – Institut des neurosciences cellulaires et intégratives INCI – Université de Strasbourg
• Dr. Sophie Tronel – Bordeaux Neurocampus
• Dr. Jean-Phillipe Guilloux – Université Paris-Sud, Université Paris Saclay

Abstract :

The physiological response to stress allows organisms to adapt. Nevertheless, when stress is sustained in time, the response becomes maladaptive and causes memory deficits and structural impairments in the hippocampus, a key region for memory. However, the mechanisms underlying these deleterious effects are not fully understood. Stress induces the release of a neuropeptide named Nociceptin/OrphaninFQ which acts on its receptors called NOP and possesses amnesic properties. We thus hypothesized that it could be a mediator of these negative effects of chronic stress.

We demonstrated that NOP receptor blockade restores long-term memory performance in mice models of chronic stress. Furthermore, we revealed a protective effect of NOP blockade on the alterations of hippocampus structure due to chronic stress.

This work paves the way for the development of new strategies using the NOP receptor as a pharmacological target for the treatment of memory deficits in stress-related disorders.

PhD defense in english

Zoom link : https://univ-tlse3-fr.zoom.us/j/92029196377?pwd=aGY4aTRDV2lSYUw3MFp6cmM4amZWZz09

Team : REMEMBER

Supervisors : Lionel DAHAN (CRCA-CBI) / Luc VALTON (CerCo)

Committee members :

• Pr. Marc Dhenain, Rapporteur, Université Paris-Saclay, Fontenay-aux-Roses
• Dr. Pierre-Pascal Lenck-Santini, Rapporteur, Aix-Marseille Université, Marseille
• Dr. Laure Peter-Derex, Rapporteure, Université Claude Bernard Lyon 1, Lyon
• Pr. Marie Sarazin, Examinatrice, Sorbonne Université, Paris
• Pr. Maria-Eugénia Soto-Martin, Examinatrice, Université Paul Sabatier, Toulouse
• Pr. Agnès Trebuchon-Da Fonseca, Examinatrice, Aix-Marseille Université, Marseille
• Dr. Lionel Dahan, Supervisor, Université Paul Sabatier, CRCA-CBI, Team REMEMBER, Toulouse
• Dr. Luc Valton, Supervisor, CHU de Toulouse, Université Paul Sabatier, CerCo, Team DYNAMO, Toulouse

Abstract :

Previous research on murine models of Alzheimer’s disease (AD) highlighted the potential role of cerebral excitatory-inhibitory imbalance in the pathophysiology of AD. Clinical research has since reported that this imbalance may give rise to epileptic activities in as much as 50% of AD patients as well. However, these epileptic events may remain undetected due to their silent, mostly sleep-occurring nature. Using a translational approach, the aim of this thesis was to better understand the link between AD, epilepsy, sleep and memory.

First, in the framework of a preclinical study, I aimed at (I) characterizing epileptic activities over the sleep-wake cycle in the Tg2576 mouse model and (II) gaining better insight into the mechanistic underpinnings of these peculiar epileptic events. Epileptic activities were frequent, showed a predominance during REM sleep, and were locked to the phase of the theta cycle corresponding to the maximal firing probability of pyramidal cells. Furthermore, these events seemed to be counterbalanced by noradrenergic signalling via α1 adrenoreceptors during wakefulness and partially during SWS, but this effect would be lost during REM sleep when noradrenergic cells cease firing. Given the early degeneration of the locus coeruleus – the primary source of noradrenaline in the brain – in AD patients, this could have important clinical implications in the future.

In the second part of this thesis, the objective was to explore the prevalence and the distribution of epileptic activities during sleep in a cohort of 30 AD patients (selected as early AD, & without epilepsy) compared to a group of age-matched, healthy controls, and to understand the impact of these activities on cognitive decline and sleep-related memory consolidation. To this end, a full-night video-EEG combined with polysomnography was undertaken, together with a two-part neuropsychological test battery with pre-and post-sleep memory tests. These measures were combined with brain imaging (MRI), genetic analyses and the evaluation of lifestyle-related risk factors. The preliminary results with 25 participants per group that are presented hint at a risk ratio for EA occurrence five times higher in patients than controls. In contrast to the preclinical results, epileptic events are predominant during non-REM sleep. Importantly, other sleep-related deficits were also observed, including an increased percentages of superficial sleep in patients at the detriment of deeper ones and unexpectedly high prevalence of newly diagnosed Sleep Apnea Syndrome: 63% in AD patients and 50% in controls. While our sample size is currently insufficient to draw steady conclusions from our results, our preliminary models suggest a combined impact of sleep quantity, sleep apnoea, epileptic events and AD itself on memory consolidation, especially on episodic memory. We suggest that treating EAs, sleep apnoea and improving sleep quality could potentially slow down or lessen memory complaints in AD patients.

PhD defense in english

Zoom parameters :

• https://univ-tlse3-fr.zoom.us/j/94449432049?pwd=M28rNGMvRnRKR2k5dEk5ZkRLeWMzdz09
• Meeting ID: 944 4943 2049
• Passcode: 623964

Supervisor : Lionel DAHAN, REMEMBeR team

Comitee members :

• M. Lionel DAHAN, Directeur de thèse, Université Toulouse III – Paul Sabatier
• M. Antoine ADAMANTIDIS, Rapporteur, University of Bern
  
• Mme Stéphanie DAUMAS, Rapporteure, Sorbonne Université
• M. Vivien  CHEVALEYRE, Rapporteur, Institute Psychiatry And Neuroscience De Paris
  
• Mme Stéphanie TROUCHE, Examinatrice, University of Montpellier
• Mme Elisa BOUTET-ROBINET, Examinatrice, Université Toulouse III – Paul Sabatier

Abstract :

The hippocampus is the main brain structure involved in episodic memory formation. The role of the hippocampus in learning, memory and their underlying mechanisms has been studied extensively in rodents, in particular by using contextual learning.

Long-Term Potentiation (LTP) is an increase in synaptic transmission of glutamatergic afferents that lasts for hours, days or months and is thought to underlie hippocampal memory formation. It can be triggered in the hippocampus by an artificial High frequency Stimulation (HFS). This mechanism helped in deciphering memory mechanisms, showing that both memory and LTP rely firstly on phosphorylation and later on de novo protein synthesis. The link between memory and LTP was confirmed by showing that blocking LTP mechanisms hinders memory formation, and that contextual learning induces LTP in the CA1 of the hippocampus. Since LTP, just like memory, can be saturated, the nervous system cannot store every sensory input that the animal encounters. Moreover, HFS is not compatible with neuronal activity. Hence, there must be a teaching signal that would be the natural molecular trigger of LTP during learning, acting as a filter choosing the pertinent inputs to store. Dopamine is a neuromodulator that has historically been thought of as a value signal, for dopamine gets released during rewarding events. However, dopamine has later been shown to be released whenever a salient unrewarding, or even punishing, event occurs. Dopamine receptors can trigger both phosphorylation and de novo protein formation in most brain structures showing plasticity, and D1/5 Dopaminergic receptors are necessary for LTP maintenance and long-term memory. Moreover, dopaminergic stimulation in vitro can modulate synaptic transmission in CA1. Thus, we hypothesized that dopamine could act as a teaching signal.

In this work, we use behavior and electrophysiology coupled with optogenetic manipulations of midbrain dopamine afferents and pharmacology inhibition of D1/5 dopaminergic receptors in order to study the role of dopamine as a teaching signal triggering LTP so that pertinent sensory inputs get stored. Using electrophysiology, we show that coupling optogenetic stimulations of midbrain dopamine with glutamatergic inputs in CA1 induces a progressive LTP that reaches its plateau 90 minutes after the pairing. This LTP endures at least 5 hours, is dependent on D1/5 receptors and partially occludes HFS-triggered LTP. Then, using contextual fear conditioning coupled with auditory cue conditioning we show that intraperitoneal injection of D1/5 receptor inhibitor, SHC23390, hinders both contextual and cue fear memories. Alternatively, intra-hippocampal infusion of SCH23390 blocks contextual memory but preserves cue fear memory intact. These results allowed us to conclude that hippocampal D1/5 receptors are necessary for contextual fear memories and in another brain structure for associative fear memories. Finally, we use a variation of contextual fear conditioning called contextual pre-exposure facilitation effect, which separates contextual learning from fear conditioning since the animal in this task learns each of them on two consecutive days. This allows studying dopamine as a teaching signal without the interference of any value inputs. We show that mice require between 2-8 minutes to encode contextual information. Furthermore, we show that D1/5 receptors are necessary for contextual and fear learning. Finally, we show that optogenetic stimulation of dopaminergic axons in the hippocampus promotes contextual learning and, conversely, their inhibition hinders contextual learning.

This work allows us to conclude that the dopaminergic pathway from the midbrain to the hippocampus has all the characteristics of a teaching signal, namely, triggering LTP on co-activated sensory inputs promoting the storage of contextual information in the hippocampus without the need for any value information.

Defense in english

Zoom Link: https://univ-tlse3-fr.zoom.us/j/93805472327?pwd=ZXFwaFVISENObXRlV0JXMFlReXM3UT09

• ID : 938 0547 2327
• Code secret : 073857

Supervisors : Martin Giurfa and Aurore Avarguès-Weber

Committee members :

• Ludovic Dickel
• Elisa Frasnelli
• Stéphane Viollet
• Aurore Avarguès-Weber
• Martin Giurfa

Abstract :

Equipped with a brain smaller than one cubic millimeter and containing ~950,000 neurons, honeybees display a rich behavioral repertoire, among which appetitive learning and memory play a fundamental role in the context of foraging activities. Besides elemental forms of learning, where bees learn specific association between environmental features, bees also master different forms of non-elemental learning, both in the visual and in the olfactory domain, including categorization, contextual learning and rule abstraction. These characteristics make them an ideal model for the study of visual learning and to explore the neural mechanisms underlying their learning abilities. In order to access the working brain of a bee during a visual learning tasks the insect needs to be immobilized. Hence, virtual reality (VR) setups have been developed to allow bees to behave within a virtual world, while remaining stationary within the real world. During my phd, I developed a flexible and open source 3D VR software to study visual learning, and used it to improve existing conditioning protocols in a virtual environment and to investigate the neural mechanism of visual learning.

Investigating the influence of optic flow on associative color learning I found that increased motion cues from the background impaired the bees’ performance. Which lead me to identify issues that may affect decision-making in VR landscapes, which require specific control by experimenters.

By means of the VR setup, I induced visual learning in tethered bees and quantified immediate early gene (IEG) expression in specific areas of their brain to detect regions involved in visual learning. In particular, I focused on kakusei, Hr38 and Egr1, three IEGs that have been related to bee foraging and orientation and thus may also be relevant when making appetitive visual association. This analysis suggests that the mushroom bodies are involved in associative color learning.

Finally, I explored the possibility of using the VR on other insect models and performed differential conditioning on bumblebees. Interestingly, not only bumblebees are able to solve this cognitive task as well as the honeybees, but they also engage more with the virtual environment, leading to a lower ratio of discarded individuals. These results indicate that the VR protocols I have established through the course of this PhD may be applied to other insects, and that the bumblebee is a good candidate for the study of visual learning under VR conditions.