Cellular & Molecular Imaging

This dataset comprises the raw data associated with Wong et al., EMBO J., 2024. "Regulation of centrosome size by the cell cycle oscillator in Drosophila embryos". This dataset (approximately 900Gb in total) tracks the behavior of individual centrosomes (labelled with different centrosome markers) in early Drosophila embryos during nuclear cycles 11-13 under various conditions. The datasets are organised to align with each individual Figure shown in the paper. The paper abstract is: Mitotic centrosomes assemble when centrioles recruit large amounts of pericentriolar material (PCM) around themselves. In early C. elegans embryos, mitotic centrosome size appears to be set by the limiting amount of a key component. In Drosophila syncytial embryos, thousands of mitotic centrosomes are assembled as the embryo proceeds through 13 rounds of rapid nuclear division, driven by a core cell-cycle oscillator. These divisions slow during nuclear cycles 11-13, and we find that centrosomes respond by reciprocally decreasing their growth rate, but increasing their growth period—so that they grow to a relatively consistent size at each cycle. At the start of each cycle, moderate CCO activity initially promotes centrosome growth, in part by stimulating Polo/PLK1 recruitment to centrosomes. Later in each cycle, high CCO activity inhibits centrosome growth by suppressing the centrosomal recruitment and/or maintenance of centrosome proteins. Thus, in fly embryos, mitotic centrosome size appears to be regulated predominantly by the core cell-cycle oscillator, rather than by the depletion of a limiting component.

This dataset comprises the raw data associated with Wong et al., EMBO J., 2024. "Regulation of centrosome size by the cell cycle oscillator in Drosophila embryos". This dataset (approximately 900Gb in total) tracks the behavior of individual centrosomes (labelled with different centrosome markers) in early Drosophila embryos during nuclear cycles 11-13 under various conditions. The datasets are organised to align with each individual Figure shown in the paper. The paper abstract is: Mitotic centrosomes assemble when centrioles recruit large amounts of pericentriolar material (PCM) around themselves. In early C. elegans embryos, mitotic centrosome size appears to be set by the limiting amount of a key component. In Drosophila syncytial embryos, thousands of mitotic centrosomes are assembled as the embryo proceeds through 13 rounds of rapid nuclear division, driven by a core cell-cycle oscillator. These divisions slow during nuclear cycles 11-13, and we find that centrosomes respond by reciprocally decreasing their growth rate, but increasing their growth period—so that they grow to a relatively consistent size at each cycle. At the start of each cycle, moderate CCO activity initially promotes centrosome growth, in part by stimulating Polo/PLK1 recruitment to centrosomes. Later in each cycle, high CCO activity inhibits centrosome growth by suppressing the centrosomal recruitment and/or maintenance of centrosome proteins. Thus, in fly embryos, mitotic centrosome size appears to be regulated predominantly by the core cell-cycle oscillator, rather than by the depletion of a limiting component.

This dataset comprises the raw data associated with Wong et al., EMBO J., 2024. "Regulation of centrosome size by the cell cycle oscillator in Drosophila embryos". This dataset (approximately 900Gb in total) tracks the behavior of individual centrosomes (labelled with different centrosome markers) in early Drosophila embryos during nuclear cycles 11-13 under various conditions. The datasets are organised to align with each individual Figure shown in the paper. The paper abstract is: Mitotic centrosomes assemble when centrioles recruit large amounts of pericentriolar material (PCM) around themselves. In early C. elegans embryos, mitotic centrosome size appears to be set by the limiting amount of a key component. In Drosophila syncytial embryos, thousands of mitotic centrosomes are assembled as the embryo proceeds through 13 rounds of rapid nuclear division, driven by a core cell-cycle oscillator. These divisions slow during nuclear cycles 11-13, and we find that centrosomes respond by reciprocally decreasing their growth rate, but increasing their growth period—so that they grow to a relatively consistent size at each cycle. At the start of each cycle, moderate CCO activity initially promotes centrosome growth, in part by stimulating Polo/PLK1 recruitment to centrosomes. Later in each cycle, high CCO activity inhibits centrosome growth by suppressing the centrosomal recruitment and/or maintenance of centrosome proteins. Thus, in fly embryos, mitotic centrosome size appears to be regulated predominantly by the core cell-cycle oscillator, rather than by the depletion of a limiting component.

The cell cycle is a fundamental process in cell biology. It consists of four phases: G1, S, G2, and M, each with distinct transcriptional programs. Its study requires a robust method to determine the cell cycle phase for a given cell. In imaging, the use of PIP-FUCCI (Fluorescent Ubiquitination-based Cell Cycle Indicator) fluorescent markers enables this characterisation, by providing images whose signal intensity is strongly correlated with the cell cycle phase. However, such markers hinder the use of other fluorescent markers that may be necessary to study the presence and/or spatial distribution of other proteins or RNAs dependent on the cell cycle phase. To address the limitations of PIP-FUCCI fluorescence, the cell cycle phase may instead be determined using an alternative fluorescent marker, such as a DNA marker like SiR-DNA. This dataset includes 982,332 images of HeLa nuclei, featuring three channels: two for PIP-FUCCI and one for SiR-DNA. Additionally, binary segmentation masks for nuclei are provided. Furthermore, 636,304 images are labeled into one of three classes — G1, S, or G2/M — enabling direct evaluation of cell cycle phase prediction models based on SiR-DNA images. If you found our work useful, please consider citing: @article{bonte2025deep, title={A Deep Learning approach for time-consistent cell cycle phase prediction from microscopy data}, author={Bonte, Thomas and Pourcelot, Oriane and Safieddine, Adham and Slimani, Floric and Mueller, Florian and Weil, Dominique and Bertrand, Edouard and Walter, Thomas}, journal={PLOS Computational Biology}, volume={21}, number={12}, pages={e1013800}, year={2025}, publisher={Public Library of Science San Francisco, CA USA} }

This dataset contains comprehensive image-based morphological profiling data from a systematic drug repurposing screen for host-directed antivirals. The repository includes three major study components. The primary screen employed high-content Cell Painting analysis combined with viral immunostaining of 5,275 repurposable compounds screened against SARS-CoV-2 infection in Vero E6 monkey kidney cells. This initial morphological profiling identified compounds that reversed the infected cell phenotype towards non-infected controls. Hit validation was performed through dose-response morphological profiling experiments by Cell Painting and antibody staining in human A549 lung epithelial cells engineered to express the ACE2 receptor, confirming antiviral activity in a human-relevant cell model. A counter screen assessed compound antiviral activity that could be attributed to drug-induced phospholipidosis (DIPL) instead of specific engagement of viral or host targets. This component includes high-content data from human A549-ACE2 cells stained with HCS LipidTOX Green Phospholipidosis Detection Reagent, CellTracker Deep Red Dye, and Hoechst 33342. The dataset includes raw images and extracted morphological features. This resource enables identification of host-targeting antivirals with potential broad-spectrum activity and supports future pandemic preparedness efforts.

The cell cycle is a fundamental process in cell biology. It consists of four phases: G1, S, G2, and M, each with distinct transcriptional programs. Its study requires a robust method to determine the cell cycle phase for a given cell. In imaging, the use of PIP-FUCCI (Fluorescent Ubiquitination-based Cell Cycle Indicator) fluorescent markers enables this characterisation, by providing images whose signal intensity is strongly correlated with the cell cycle phase. However, such markers hinder the use of other fluorescent markers that may be necessary to study the presence and/or spatial distribution of other proteins or RNAs dependent on the cell cycle phase. To address the limitations of PIP-FUCCI fluorescence, the cell cycle phase may instead be determined using an alternative fluorescent marker, such as a DNA marker like SiR-DNA. This dataset includes 982,332 images of HeLa nuclei, featuring three channels: two for PIP-FUCCI and one for SiR-DNA. Additionally, binary segmentation masks for nuclei are provided. Furthermore, 636,304 images are labeled into one of three classes — G1, S, or G2/M — enabling direct evaluation of cell cycle phase prediction models based on SiR-DNA images. If you found our work useful, please consider citing: @article{bonte2025deep, title={A Deep Learning approach for time-consistent cell cycle phase prediction from microscopy data}, author={Bonte, Thomas and Pourcelot, Oriane and Safieddine, Adham and Slimani, Floric and Mueller, Florian and Weil, Dominique and Bertrand, Edouard and Walter, Thomas}, journal={PLOS Computational Biology}, volume={21}, number={12}, pages={e1013800}, year={2025}, publisher={Public Library of Science San Francisco, CA USA} }

Indocyanine Green (ICG) has seen widespread use in the operating room as a fluorescent vascular imaging agent. However, its rapid vascular clearance often necessitates redosing, resulting in procedural delays, greater background, and potential adverse physiological effects. Here, we present Chromenylium Green (ChromG), a novel fluorescent tracer exhibiting excellent visualization of vasculature in mice with greater contrast and seven times the vascular half-life relative to ICG. ChromG allows for high-resolution imaging of vasculature at greater depths, enabling whole-body 3D vascular reconstruction with up to 74% connectivity. As a proof of concept of ChromG’s utility in disease contexts, we demonstrate the ability to visualize saphenous artery stenoses above the clinically accepted contrast limit for eight times longer than ICG. With its excellent biocompatibility profile and compatibility with current clinical imaging technologies, ChromG is a promising fluorescent probe for prolonged, high-resolution intraoperative vascular imaging.

Images show the effects of purified retromer on oriented membrane tubes and the effects of mutant versions of retromer in living yeast and mammalian cells

This dataset includes microscopy and sequence data related to the study and used in the figures. It includes light and electron microscopy images. The sample preparation, image acquisition and analysis protocols are described in the methods.

Fluorescence imaging across the near-infrared (NIR, 700–1000 nm) and shortwave infrared (SWIR, 1000–2000 nm) regions offers significant advantages for biomedical applications, yet photophysical enhancements achieved with NIR and SWIR chromophores observed in solution often fail to translate to complex biological environments. Fluorous-soluble fluorophores, fluorofluorophores, face additional challenges, exhibiting poor brightness and photostability when dissolved in perfluorocarbons (PFCs) due to unfavorable interactions with the fluorous phase. Here, we report counterion exchange as a strategy to enhance the photophysical properties of two heptamethine cyanine fluorofluorophore for NIR and SWIR imaging. Exchanging the small chloride counterion with a large, fluorinated aryl borate counterions significantly improved the brightness (10-fold) and photostability (57-fold) in PFCs. These enhancements were successfully translated across multiple biological systems from macrophage cells to NIR imaging zebrafish retinal tissue and finally to SWIR imaging in mice. These results demonstrate that strategic counterion modification provides a straightforward approach to optimize fluorofluorophores, with solution-phase improvements that translate to in vivo NIR and SWIR imaging.