Pyrotechnic devices that generate smoke are used in many fields such as civil or military ones. People in contact with smokes are exposed by respiratory route or by eye and skin contact and those smokes may lead to adverse effects. Currently, the assessment of toxicological hazard implies the use of animal testing which, in addition to the ethical problems, is time and costs consuming. This is a major issue for manufacturers since the development of pyrotechnic formulations involves an important R&D investment to obtain effective and safe products for users.
Considering current societal concerns associated to animal testing, efforts have been engaged to develop new alternative methods to estimate the level of toxicity related to the acute exposure of a pyrotechnic composition by progressively replacing in vivo experiments by in vitro tests. This alternative approach will also favor Safer By Design products by allowing toxicity evaluations earlier in the development of new pyrotechnic formulations.
In a first step for this purpose, various cell models were exposed at the air liquid interface (ALI) to smokes generated from known pyrotechnic compositions. Three different compositions with different range of toxicity based on published data were studied. Pulmonary cell lines, corneal, and epidermal primary cells were used to reproduce the main routes of exposure. In this PhD framework a greater focus on pulmonary cell line was brought. The toxicological effect was assessed by cell viability, inflammation, and oxidative stress markers. The in vitro results obtained were in accordance with literature in vivo data regarding the toxicity of these pyrotechnic compositions. It was therefore possible to discriminate between different toxicity range of the smokes.
Further in vitro and in vivo exposure correlation will still be needed as a perspective to confirm this in vitro technique as a new standard. Once the correlation will be validated, in vitro testing could be used as a reliable alternative method for the assessment of the toxicity of newly developed pyrotechnic composition.
Aya ASHRIN
​
Development of an in vitro method to assess the toxicity of pyrotechnic fume compositions
Insects use odorant receptors (ORs) to perceive odors. The project aims to understand how duplicated ORs acquire distinct response spectra following duplication. The study will use sequence data from Spodoptera species and other Lepidoptera to trace the evolutionary trajectory leading to functional divergence between OR duplicates. The research will focus on reconstructing evolutionary trajectories, identifying key amino acid substitutions, and functionally assessing ancestral ORs to understand the molecular basis of odorant recognition and the evolution of the OR gene family.
Abhinob BARUAH
​
Reconstructing the evolution of insect odorant receptors
Maize landraces are a valuable source of genetic diversity for facing climate change due to their local adaptation to various agro-climatic conditions. High-throughput pool genotyping (HPG) is a cost-effective approach to genotype maize landraces and identify promising sources of favourable alleles or populations for tolerance to abiotic stress (Arca et al., 2023). We applied this approach on a large world-wide collection of maize landraces to i) characterize its genetic structuration; ii) identify genomic regions involved in adaptation through Genome Environment Association studies (GEA) and Genome Wide Association studies (GWAS) on adaptative traits; iii) perform genomic prediction (GP) of both adaptive and agronomic traits based on a subset of landraces that are both phenotyped and genotyped.
Agustin GALARETTO
​
Genomic prediction and landscape genomics in a large maize landraces collection using high-troughput pool genotyping identifies promising sources of diversity for prebreeding
Recombinant proteins are present in many aspects of our lives. Several organisms are used for their production, among which the yeast Saccharomyces cerevisiae. Expressing a protein of interest in its non-native organism requires optimization by engineering the host organism, the promoter or the peptide-signal, which are parameters defining the secretion background. Currently, optimization of the secretion background is made with low-throughput methods and faces protein-specific bottlenecks making the task non-trivial and time consuming.
Our project aims to explore a broad diversity of protein secretion backgrounds in yeast and identify the best ones in a high throughput manner by screening pooled libraries of variants, i.e. population where each cell has a different secretion background. We want to build a growth-based selection system, meaning that variants conferring high secretion capacity grow faster and are enriched at the end of the screening process. Concretely, we engineer a genetic circuit to create self-communication, meaning that each cell senses its own recombinant protein production, triggering in response the expression of an essential gene. The sensing system is based on the hijacking of the yeast endogenous mating pathway.
Leveraging our automated bioreactor platform, we were able to establish robust autocrine-like communication and monitor growth quantitatively. The next steps consist in understanding better how the sensing subunit operates and testing the circuit in several secretion backgrounds already known. Finally, the system will be used to screen libraries of variants by deep sequencing to compute enrichment of all variants of the library after selection.
Henri GALEZ
​
Me, myself and I: engineering self-communication in yeast to improve recombinant protein production
The team has selected small thermostable artificial proteins, called αReps, that are able to block SARS-CoV-2 infection by interacting with the viral S protein. Although they presented excellent results in vitro and in vivo via intranasal treatment, they were quickly absorbed by the nasal epithelium, which lowers the antiviral protection. We linked our antivirals with a mucin-binding domain called X409, which would increase the protein intranasal residence time by targeting the mucus covering the nasal cavity. In the hamster model, this new construct (αReps-X409) not only increased residence time but also provided antiviral protection by decreasing viral load in the nasal cavity and lungs. It also presented low toxicity and immunogenicity in zebrafish larvae and mice, respectively.
Joelton ROCHA GOMES
​
A spray twice a day to keep COVID-19 away
Newly developed plant-based foods can vary in protein, but can consumers detect these differences? Here, we sought to explore whether repeated consumption of a novel plant-based protein-containing food impacts the macronutrient profile of other foods in the diet.
Healthy adult meat-eating participants consumed a portion of a novel plant-based test food daily over 5 days. In a between-group design, two levels of protein, 10% and 25%, were tested. Participants kept a digital food diary and attended an experimental session on day 1 and day 8. During the experimental sessions, participants ate the test food (preload), followed by an ad libitum buffet. They also evaluated their liking for the test food, alone or in combination with protein or carbohydrates, together with its expected satiety.
Thirty-eight (n(25%)=20) participants completed the intervention. The higher (25%) protein test food received lower liking and higher expected satiety ratings (p<0.05). Both groups declared a greater desire to eat the test food in combination with carbohydrates than with protein (p<0.001). There were no between-group differences in macronutrient intake at the buffet, but the 25% group increased their lipid and calorie intake by 14% between sessions (p<0.05). Food diaries show an 20% reduction in daily protein intake in the 25% group over 5 days (p=0.04).
In this study, participants used test foods as protein sources independently of their protein content. However, after several days of intervention, the higher-protein group reduced its daily protein intake indicating a potential compensation on physiological level.
Marjorie GOURRU
​
Effects of a novel plant-based protein on daily nutritional intake: do people adapt over time ?
French beekeepers are in demand for references and methods to structure and optimize their selection work, particularly to jointly consider different breeding goals: productivity, and resilience defined as a good general resistance to diseases and pests (primarily the mite Varroa destructor) as well as a strong food autonomy to limit the supply of syrup and candy sugar to colonies when floral resources are insufficient. The project aims at meeting this strong demand from the bee industry by developing sustainable breeding objectives and selection programs to efficiently manage beekeeping productions. The core of the project is the design of multi-criteria breeding plans for honey bees and the assessment of their potential efficiencies through the joint analysis of field data and simulated ones for theoretically designed breeding plans. The three main questions to be addressed are:
1) How should be defined the multitrait breeding goal to get the right compromise between the stakeholders’ interests for production and resilience traits?
2) How accurately can be estimated the corresponding genetic parameters according to the data and population structure?
3) Given the inaccuracy of the genetic parameters’ estimates in honeybee populations linked to uncertainty in pedigrees and complexity of colony performance, which breeding strategy should be recommended to get maximum genetic gain while limiting inbreeding?
Tristan KISTLER
​
How to design efficient multitrait selection breeding plans for honeybees ?
L'étude porte sur l'analyse numérique du métabolisme d'une bactérie soumise à du bruit. La représentation du métabolisme bactériel se fait par sa modélisation en réseaux métabolique. Ce modèle numérique est ensuite soumis à des conditions environnementales et internes incertains. Des outils de la théorie de l'information sont ensuite utilisé pour lier entre eux les différents éléments du réseaux et ainsi interpréter les niveaux de dépendance au sein du réseaux.
Arthur LEQUERTIER
​
Etude du transfert d'information dans le metabolisme bacteriel en environnement bruité
This work aims to develop a physiologically based kinetic model, linked to a mechanistic toxicodynamic model of the hypothalamic–pituitary–gonadal (HPG) axis in zebrafish (PBK-HPG model), that could link the dose to adverse effects and be used for risk assessment of endocrine disruptors (ED). It was applied to two ED-suspected azole fungicides: prochloraz (PCZ) and imazalil (IMZ), both known for their aromatase inhibition potency.
The proposed PBK-HPG model comprises twelve compartments representing diverse tissues and adapts to various exposure scenarios. It considers vitellogenin (VTG) and steroid hormones (estradiol, testosterone, 11-ketotestosterone), focusing on key aspects such as brain feedback loops, liver VTG synthesis, and gonadal steroid hormone synthesis. Model predictions were compared to original data from OECD TG229 and a time-dependent measurement of steroid concentrations.
The model accurately predicted internal concentrations of PCZ and IMZ in key organs. It faithfully replicated the HPG axis baseline physiological conditions. Simulations of PCZ and IMZ effects on the HPG axis showed good results. Overall, the model was able to reproduce all the different experimental conditions.
To conclude, our model provides valuable insights into the dynamics of the HPG axis, offering an understanding of the mechanism underlying the impact of azole fungicides on the HPG axis in zebrafish. The novel PBK-HPG model holds promise for integration into a quantitative adverse outcome pathway framework, offering a comprehensive approach for assessing the toxic effects of ED.
Tu-Ky LY
​
Modeling Azole Fungicides' Endocrine Disruption: A Physiologically Based Kinetic Model of the Hypothalamus-Pituitary-Gonads Axis (PBK-TD) in Zebrafish.
Individual animal behaviour monitoring can provide an early sign of variation in animal welfare. Artificial Intelligence and sensors are promising to predict animal behaviour automatically. This paper aims to present a pipeline using a supervised classification algorithm called ModBehav for automatically predict certain animal behaviours from accelerometer data. This pipeline is designed to be generic and applicable to different species, behaviours and types of accelerometers. It was applied on data obtained from 8 indoor-housed goats equipped with ear-mounted accelerometers. “Ruminating”, “head in the feeder”, “lying” and “standing” were continuously sampled from camera recordings for 11 hours for each goat to evaluate the model performances, using the AUC score (Area Under the Curve). AUC score plots the True Positive against the False Positive rate across various threshold values, when common metrics such as accuracy, sensitivity, specificity and F1-score use a fixed-threshold. AUC score provides a reliable metric to identify which model is better on average. For each behaviour, various filtering techniques, time-window segmentations, additional time-series data, and feature selections were tested. The best processing treatments were selected based on the highest AUC score obtained. “Ruminating”, “head in the feeder”, “lying” and “standing” were best predicted with AUC scores of 0.792, 0.806, 0.828 and 0.832, respectively. For all behaviours, a decrease in the AUC score was observed when the prediction was made with data from goats that have not been used during the training of the model.
Sarah MAUNY
​
Modbehav, a generic supervised classification algorithm to predict animal behaviour from accelerometer data
La SNCF est le 2e propriétaire foncier après l’Etat, avec notamment 88 000 hectares de dépendances vertes à entretenir chaque année. Cela entraine une nécessité de traiter la végétation présente sur ce foncier, et notamment aux abords des voies, pour des enjeux sécuritaires, financiers, législatifs et durables. De ce fait, la maitrise de la végétation est aujourd’hui le 2e poste de dépense le plus important chez SNCF Réseau, ce qui représente une multiplication du budget de 2,5, par rapport à 2015. Ceci s’explique notamment par une complexification des législations, qui a obligé un passage à des pratiques de traitement plus manuelles, et qui parallèlement a conduit à une baisse de l’efficacité de maintenance. La SNCF souhaite donc valoriser ces végétaux pour réduire ses coûts de maintenance, et s’inscrire durablement dans le territoire qu’elle traverse.
La thèse a donc pour objectif d’identifier et d’évaluer les scénarios de valorisation possibles à l’aide de méthodes économiques et sociologiques. Le poster vise à donner plus de détails sur ces objectifs, les méthodes utilisées, les premiers résultats obtenus, et les prochaines étapes.
Loup SARDIN
​
Valorisation des déchets végétaux ferroviaires en Ile-De-France
Hepatitis A virus (HAV) is an enteric virus transmitted via the fecal-oral route and the foodborne transmission rate is up to 16% (Van Cauteren, 2018). Frozen berries imported from endemic countries for HAV can lead to foodborne diseases because they are mainly eaten uncooked and not or barely washed. HAV is known to resist freezing but the impact of the thermic history (freezing and storage conditions) on its persistence is not established. About food processing, the impact of thermic history on microstructure and quality development is established for several food matrices, but few data are available concerning frozen raspberries quality. The QUALISURE project aims to study not only the persistence but also the microstructure and quality of raspberries submitted to different thermic histories.
Artificially contaminated raspberries with HAV were frozen and stored at -5°C or -18°C. Infectious viruses were quantified along the storage using impedance-based cell culture systems. In parallel, frozen raspberries microstructure were studied using X-Ray Computed micro-Tomography, and the organoleptic quality were evaluated by studying texture and drip loss. Infectious HAV remained stable over time on raspberries stored at -18°C for 4 months but decreased as no infectious viruses could be detected after 4 months of storage at -5°C. Minimal changes of raspberries microstructure and quality were observed with -18°C storage unlike -5°C storage, which led to severe damages.
These results taken together, show how storage temperature can affect microstructural and macroscopical quality parameters, as the persistence of infectious viruses and provide new perspectives in multiscale food research.
Gwenaëlle VERBRUGGHE
​
From microstructure development to quality changes and viral risk: multiscale analysis of frozen raspberries.