Biology subjects

Joly, A.

Publications and source records attributed to Joly, A..

12 recordsLinked to original sources

bioRxiv2026

Text guidance is powerful but prompt-sensitive for weakly-supervised leaf symptom segmentation

Accurate segmentation of plant disease symptoms is essential for crop monitoring and phenotyping, yet it typically requires costly pixel-level annotations. Weakly supervised semantic segmentation (WSSS) alleviates this burden using image-level labels, but its performance depends on the quality of spatial priors such as class activation maps (CAMs). We investigate whether text-guided segmentation with the Segment Anything Model 3 (SAM3) can serve as an alternative weak supervision signal. Three pseudo-mask generation strategies are compared: (i) CAMs refined with SAM or SAM3, (ii) zero-shot text-guided SAM3, and (iii) a hybrid approach combining weak spatial cues with text prompts. The resulting pseudo-masks are used to train a DeepLabV3 model. Text guidance alone matches or outperforms conventional WSSS, achieving up to 0.46 IoU without spatial supervision and 0.61 IoU on a public dataset, although performance is sensitive to text prompt formulation. The hybrid strategy improves robustness, reaching 0.50 IoU on the primary dataset and 0.58 IoU on the additional dataset while reducing prompt sensitivity. Overall, text guidance is a promising alternative to conventional weak supervision, while hybrid approaches provide a more robust solution for plant disease segmentation.

plant biology↗

bioRxiv2026

Coherent Structured Illumination Microscopy with Enhanced Optical Sectioning

Coherent structured illumination microscopy (c-SIM) is a synthetic aperture optical technique for sub-diffraction limit imaging that extends the utility of traditional SIM to non-fluorescent samples. Here, we present a complementary 5-beam implementation of c-SIM that provides enhanced optical sectioning compared to conventional quadrupolar illumination. Since our approach detects intensity images due to coherent light scattering, it avoids the complications associated with detecting complex fields. Through comparative measurements on calibration samples and live microalgae, we show that 5-beam c-SIM effectively suppresses coherent defocus effects, improving image quality while simultaneously providing a 2-fold lateral resolution improvement.

bioengineering↗

bioRxiv2025

Liver FGF21 mediates sex-specific adaptation to juvenile protein malnutrition

Dietary protein sufficiency during childhood is essential for healthy growth and tissue development. Chronic protein deficiency leads to stunting, height-for-age more than two standard deviations below the median, and affects 149 million children under five (https://www.who.int/publications/i/item/9789240073791). Linear growth is governed by the somatotropic axis, wherein pituitary growth hormone (GH) stimulates hepatic insulin-like growth factor 1 (IGF-1) to promote bone elongation1. Although boys are consistently more stunted than girls across populations (ranging from 18% to 45%)2,3, mechanisms underlying this apparent resilience remain unclear. Rodent studies suggest that the protein-to-carbohydrate ratio modulates life-history traits such as metabolism, lifespan, and reproduction, with evidence of sex-dependent sensitivity4,5. Whether such dimorphism reflects coordinated developmental adaptations in response to protein scarcity, particularly in females, has not been fully explored. Here, we show that juvenile dietary protein restriction causes sexually dimorphic adaptations with pronounced stunting in males and delayed reproductive maturation in females. Using liver-specific deletion, we identify fibroblast growth factor 21 (FGF21) as a sex-specific hepatic regulator of female somatic growth and reproductive development, revealing a dimorphic endocrine adaptation to early-life protein scarcity. Our findings position FGF21 as a critical physiological checkpoint coordinating growth and reproductive timing in response to nutritional stress.

physiology↗

bioRxiv2025

Chronic dietary ingestion of Bacillus thuringiensis spores promotes intestinal inflammation and aging in non-target adult Drosophila

Bacillus thuringiensis (Bt), a spore-forming Gram-positive bacterium, is the leading microbial insecticide used in both organic and conventional agriculture to fight lepidopteran larvae. Bt insecticides, which consist of a mix of Bt spores and crystals of entomopathogenic toxins (Cry toxins), kill target pest larvae rapidly after ingestion by destroying their intestinal epithelium. The potential adverse effects of chronic consumption of Bt spores by non-target organisms have not been thoroughly investigated. In this study, using adult Drosophila melanogaster, a non-target dipteran organism of Bt insecticides, we show that chronic ingestion of agricultural doses of spores in the diet reduces lifespan. Our results demonstrate that Bt spore ingestion affects gut morphology, promotes dysplasia, alters septate junctions and enhances epithelial permeability. In addition, we observed increased levels of inflammatory signaling pathways and reactive oxygen species. Taken together, our results indicate that chronic consumption of Bt spores promotes inflammation and oxidative stress, leading to premature aging of the gut and early lethality in Drosophila. Extended to non-target insects, which account for 85% of animal biodiversity, our study suggests that highly persistent Bt spores may have unintended long-term effects on the environment.

physiology↗

bioRxiv2025

A tumor-suppressive role of the PRC1 Polycomb epigenetic complex in the maintenance of adult Drosophila intestinal stem cell identity

Chromatin modulators, like Polycomb group proteins, are key epigenetic regulators of gene expression and are frequently mutated in cancers. In adult stem cells, epigenetic regulation maintains their identity and controls their differentiation during homeostasis or aging, but its direct role in tumorigenesis remains unclear. Here we developed a novel tumor model in Drosophila by exploring the function of Polycomb Repressive Complex 1 (PRC1) in adult intestinal stem cells (ISCs). Disrupting core PRC1 components in ISCs induces the formation of small cell clusters devoid of intestinal markers, a novel phenotype linked to premature mortality under stress. These clusters exhibit neoplastic characteristics such as overproliferation and continuous growth in serial transplantations, leading to their designation as tumor-initiating intestinal cells (TIICs). While JAK/STAT signaling contributes to TIIC growth, the NF-{kappa}B-related Toll/Imd immune pathways restrict their expansion independently of cell death. Altogether, our results highlight PRC1 as an epigenetic tumor suppressor in adult stem cells. AUTHOR SUMMARYOur research explores how stem cells in the adult intestine stay healthy and avoid becoming cancerous. We focused on a group of proteins called Polycomb Repressive Complex 1 (PRC1), which help regulate which genes are turned off in a cell. While these proteins are known to play important roles during development and cancer prevention, their function in the adult intestine has been less clear. Using the fruit fly Drosophila, we discovered that when PRC1 function is lost in intestinal stem cells, abnormal clusters of cells begin to form. These clusters grow uncontrollably, lose their normal identity, and can keep growing when transplanted--traits that are typical of cancer. Interestingly, we also found that parts of the immune system, specifically NF-{kappa}B-related pathways, can act inside these tumor cells to limit their growth--revealing a protective role that hasnt been seen before in the adult gut. This work provides new insights into how epigenetic regulation and immune signaling work together to keep stem cells from turning cancerous. It opens up new possibilities for understanding how cancers begin and how the body may naturally resist them, even at the level of the tissue.

cancer biology↗

bioRxiv2025

Stem cell-specific NF-kB is required for stem cell survival and epithelial regeneration upon intestinal damage.

Immune signals coordinate the repair of damaged epithelia by intestinal stem cells. However, it is unclear if immune pathways act autonomously within the stem cell to direct the damage response pathway. We consider this an important question, as stem cell dynamics are essential for formation and maintenance of the entire epithelium. We used Drosophila to determine the impact of stem cell-specific loss of NF-{kappa}B on tissue regeneration upon chemical injury. We found that loss of NF-{kappa}B enhanced cell death, impaired enterocyte renewal and increased mortality. Mechanistically, we showed that inhibition of stem cell apoptosis is essential for NF-{kappa}B-dependent maintenance of cell viability and tissue repair. Combined, our data demonstrate that stem cell-intrinsic NF-{kappa}B activity is essential for an orderly repair of damaged intestinal epithelia.

immunology↗

bioRxiv2024

Protein restriction associated with high fat induces metabolic dysregulation without obesity in juvenile mice

Dysregulation of energy metabolism, including hyperglycemia, insulin resistance and fatty liver have been reported in a substantial proportion of lean children. However, non-obese murine models recapitulating these features are lacking to study the mechanisms underlying the development of metabolic dysregulations in lean children. Here, we develop a model of diet-induced metabolic dysfunction without obesity in juvenile mice by feeding male and female mice a diet reflecting Western nutritional intake combined with protein restriction (mWD) during 5 weeks after weaning. mWD-fed mice (33% fat, 8% protein) do not exhibit significant weight gain and have moderate increase in adiposity compared to control mice (16% fat, 20% protein). After 3 weeks of mWD, juvenile mice have impaired glucose metabolism including hyperglycemia, insulin resistance and glucose intolerance. mWD also triggers hepatic metabolism alterations, as shown by the development of simple liver steatosis. Both male and female mice fed with mWD displayed metabolic dysregulation, which a probiotic treatment with Lactiplantibacillus plantarum WJL failed to improve. Overall, mWD-fed mice appear to be a good preclinical model to study the development of diet-induced metabolic dysfunction without obesity in juveniles.

physiology↗

bioRxiv2024

Semaglutide interferes with postnatal growth in a diet-dependent manner

GLP-1 receptor agonists, including semaglutide, are increasingly used to treat obesity--and to a lesser extent, type 2 diabetes--in children and adolescents. Despite their therapeutic promise, the impact of semaglutide on key endocrine and metabolic developmental processes during this period is not yet fully understood. To address the limited understanding of GLP-1 receptor agonists during development, we investigated the effects of chronic semaglutide treatment on growth and metabolism of weanling male mice. In juvenile mice displaying a normal growth pattern, semaglutide lowered blood glucose without affecting food intake or weight gain. In contrast, chronic semaglutide treatment in growth-stunted juveniles exacerbated linear growth impairment, despite preserving glucose-lowering effects. Transcriptomic analysis of the liver, complemented by gene expression studies in other metabolically active tissues, suggested altered energy expenditure, indicating a complex interaction between semaglutide, energy metabolism, and growth during juvenile development. These findings underscore the importance of developmental context when evaluating the safety and efficacy of GLP-1 receptor agonists in pediatric populations.

physiology↗

bioRxiv2023

Predicting species distributions in the open oceans with convolutional neural networks.

As biodiversity plummets due to anthropogenic disturbances, the conservation of oceanic species is made harder by limited knowledge of their distributions and migrations. Indeed, tracking species distributions in the open ocean is particularly challenging due to the scarcity of observations and the complex and variable nature of the ocean system. In this study, we propose a new method that leverages deep learning, specifically convolutional neural networks (CNNs), to capture spatial features of environmental variables. This novelty eliminates the need to predefine these features before modelling and creates opportunities to discover unexpected correlations. Our aim is to present the results of the first trial of this method in the open ocean, discuss limitations and provide feedback for future improvements or adjustments. In this case study, we considered 38 taxa comprising pelagic fishes, elasmobranchs, marine mammals, marine turtles and birds. We trained a model to predict probabilities from the environmental conditions at any specific point in space and time, using species occurrence data from the Global Biodiversity Information Facility (GBIF) and environmental data from various sources. These variables included sea surface temperature, chlorophyll concentration, salinity and fifteen others. During the testing phase, the model was applied to environmental data at locations where species occurrences were recorded. The classifier accurately predicted the observed taxon as the most likely taxon in 69% of cases and included the observed taxon among the top three most likely predictions in 89% of cases. These findings show the adequacy of deep learning for species distribution modelling in the open ocean. Additionally, this purely correlative model was then analysed with explicability tools to understand which variables had an influence on the models predictions. While variable importance was species-dependent, we identified finite-size Lyapunov exponents (FSLEs), sea surface temperature, pH and salinity as the most influential variables, in that order. These insights can prove valuable for future species-specific ecology studies.

ecology↗

bioRxiv2022

Talin and kindlin cooperate to control the density of integrin clusters.

Focal adhesions are important mechanosensitive structures, composed of transmembrane integrins, linking the extracellular matrix to the actomyosin cytoskeleton, via cytoplasmic proteins. Cellular adhesion to the extracellular matrix depends on the activation of integrins by intracellular mechanisms. Talin and kindlin are major activators of integrins that are recruited to the inner membrane and bind to {beta}-integrin cytoplasmic tails. Many studies showed the importance of integrin activation and clustering and how the organization of extracellular ligands guides the nanoscale organization of adhesion complexes. However, the roles of talin and kindlin in this process are poorly understood. To determine the contribution of talin, kindlin, lipids and actomyosin in integrin clustering, we performed experiments using a biomimetic in vitro system, made of Giant Unilamellar Vesicles, containing transmembrane integrins, on which purified talin, kindlin, and actomyosin assemble. Here we first show that talin and kindlin individually have the ability to cluster integrins. When added together, talin and kindlin synergize to induce the formation of larger integrin clusters containing the three proteins. Comparison of protein density in the talin-integrin, kindlin-integrin, and talin-kindlin-integrin clusters reveals that kindlin increases talin and integrin density, whereas talin does not affect kindlin and integrin density. Finally, kindlin significantly enhances the segregation of talin-integrin clusters induced by actomyosin contractility, suggesting that it increases the coupling of these clusters to the actin cytoskeleton. Our study unambiguously demonstrates how kindlin and talin cooperate to induce integrin clustering, which is a major parameter for cell adhesion.

biochemistry↗

bioRxiv2022

Arginine-dependent hypusination of the eukaryotic translation initiation factor (eIF)5A drives erythroid lineage differentiation

Metabolic programs contribute to hematopoietic stem and progenitor cell (HSPC) fate but it is not known whether the metabolic regulation of protein synthesis controls HSPC differentiation. We discovered that SLC7A1/CAT1-dependent arginine uptake and its catabolism to spermidine control the erythroid specification of HSPCs via activation of eukaryotic translation initiation factor 5A (eIF5A). eIF5A activity is dependent on the metabolism of spermidine to hypusine and inhibiting hypusine synthesis abrogates erythropoiesis and diverts EPO-stimulated HSPCs to a myeloid fate. Proteomic profiling reveals mitochondrial translation to be a critical target of hypusinated eIF5A and induction of mitochondrial function partially rescues erythropoiesis in the absence of hypusine. Within the hypusine network, ribosomal proteins are highly enriched and we identify defective eIF5A hypusination in erythroid pathologies caused by abnormal ribosome biogenesis. Thus, eIF5A-dependent protein synthesis is critical in the branching of erythro-myeloid differentiation and attenuated eIF5A activity characterizes ribosomal protein-linked disorders of ineffective erythropoiesis.

physiology↗

bioRxiv2020

A new method for counting reproductive structures in digitized herbarium specimens using Mask R-CNN

Phenology-the timing of life-history events-is a key trait for understanding responses of organisms to climate. The digitization and online mobilization of herbarium specimens is rapidly advancing our understanding of plant phenological response to climate and climatic change. The current practice of manually harvesting data from individual specimens, however, greatly restricts our ability to scale-up data collection. Recent investigations have demonstrated that machine-learning approaches can facilitate this effort. However, present attempts have focused largely on simplistic binary coding of reproductive phenology (e.g., presence/absence of flowers). Here, we use crowd-sourced phenological data of buds, flowers, and fruits from > 3000 specimens of six common wildflower species of the eastern United States (Anemone canadensis L., A. hepatica L., A. quinquefolia L., Trillium erectum L., T. grandiflorum (Michx.) Salisb., and T. undulatum Wild.) to train models using Mask R-CNN to segment and count phenological features. A single global model was able to automate the binary coding of each of the three reproductive stages with > 87% accuracy. We also successfully estimated the relative abundance of each reproductive structure on a specimen with [≥] 90% accuracy. Precise counting of features was also successful, but accuracy varied with phenological stage and taxon. Specifically, counting flowers was significantly less accurate than buds or fruits likely due to their morphological variability on pressed specimens. Moreover, our Mask R-CNN model provided more reliable data than non-expert crowd-sourcers but not botanical experts, highlighting the importance of high-quality human training data. Finally, we also demonstrated the transferability of our model to automated phenophase detection and counting of the three Trillium species, which have large and conspicuously-shaped reproductive organs. These results highlight the promise of our two-phase crowd-sourcing and machine-learning pipeline to segment and count reproductive features of herbarium specimens, thus providing high-quality data with which to investigate plant response to ongoing climatic change.

ecology↗