Biology subjects

Di Sante, M.

Publications and source records attributed to Di Sante, M..

5 recordsLinked to original sources

bioRxiv2025

LigHTS: Massively Parallel Biomimetic Photo-Functionalization for Imaging-Based Ultra-High-Throughput Screening

Imaging-based ultra-high-throughput screening (UHTS) in pharma and biotech still runs on 384/1536-well plates whose stiff, flat substrates limit biological fidelity and screening efficiency. Highly biomimetic organs-on-chips and organoids improve relevance but lack reproducibility and plate-scale throughput. Biomimetic hydrogel scaffolds can be produced at scale through photopolymerization, which yet uses focused optics to define micrometer-resolved geometries, constraining scalability. To address the technical challenge of truly scalable biomimetic substrates featuring anisotropies, this study presents LigHTS, an all-optical, in-well method that replaces focused with collimated illumination to photofabricate structured hydrogels in standard 384/1536-well plates. Adding food dye tartrazine to gelatin-methacrylate (GelMA) solutions increases hydrogel thickness sensitivity to UV dose by [~]10x, allowing uniform control of film thickness without lenses. Entire plates are functionalized in parallel with soft hydrogels ([~]1-10 kPa) whose thickness is tunable from 10 to 100 {micro}m. Simultaneously, simply interposing film photomasks encoding anisotropies enables orthogonal control of thickness and topography at UHTS throughput. Biological effect is demonstrated with mechanosensitive HT1080 cells, which display stiffness- and topography-dependent spreading and contact-guided migration on LigHTS-produced grooved substrates. Geometric uniformity across the plate (coefficient of variation <20%) meets HTS reproducibility standards, providing a readily available solution with enhanced biomimicry for imaging-based UTHS pipelines.

bioengineering↗

bioRxiv2025

Vertically Integrated System for Tracking and Assessing cell-cycle aware phenotypes under confinement

Quantitative cell biology often studies migration and the cell-cycle (CC) in separate assays, limiting mechanistic insights, particularly under geometric confinement. Here, we introduce a vertically integrated platform for simultaneously tracking single-cell migration and assessing CC under confinement. Our system integrates cell engineering via multiplexed sensors for cell-cycle, actin, and tubulin, as well as photopatterned engineered extracellular matrix (ECM) islands of defined sizes. It also features an automated, high-throughput pattern-aware imaging pipeline (Fab2Mic) that enables on-pattern, joint migration-CC assessment in the same live cells. Since the local microenvironment plays a critical role in metastasis by constraining cell behaviors within spatial boundaries, we used an HT1080 fibrosarcoma model as an illustrative case. Where static phenotyping yielded 40% G1 and 60% S/G2/M, with larger cell areas and tubulin spread in the S/G2/M phase, dynamic phenotyping via live-cell imaging confirmed CC-linked motility, with faster instantaneous velocities in G1, exemplifying the CC-migration correlations. These phenotypes were modulated by the spatial confinement imposed by the engineered ECM islands. Stronger confinement reduced cell area and tubulin spread and increased the frequency of abnormal CC events, particularly Long G1 states on smaller engineered ECM islands. It also induced a confinement-specific S/G2/M-G1 mitotic slippage, observed only under our confined conditions. Together, this vertically integrated system suggests that confinement may continuously tune migration-CC coupling and provides a deployable pipeline for CC-aware mechanobiology and screening. Moreover, we stress how dynamic imaging provides access to variables that are difficult or impossible to infer from static snapshots, including velocity and CC timing.

bioengineering↗

bioRxiv2025

MicroSplit: Semantic Unmixing of Fluorescent Microscopy Data

Fluorescence microscopy, a key driver for progress in the life sciences, faces limitations due to the microscopes optics, fluorophore chemistry, and photon exposure limits, necessitating trade-offs in imaging speed, resolution, and depth. Here, we introduce Micro[S]plit, a computational multiplexing technique based on deep learning that allows multiple cellular structures to be imaged in a single fluorescent channel and then unmix them by computational means, allowing faster imaging and reduced photon exposure. We show that Micro[S]plit efficiently separates up to four superimposed noisy structures into distinct denoised fluorescent image channels. Furthermore, using Variational Splitting Encoder-Decoder (VSE) networks, our approach can sample diverse predictions from a trained posterior of solutions. The diversity of these samples scales with the uncertainty in a given input, allowing us to estimate the true prediction errors by computing the variability between posterior samples. We demonstrate the robustness of Micro[S]plit networks, which are trained for each splitting task at hand, across various datasets and noise levels and show its utility to image more, to image faster, and to improve downstream analysis. We provide Micro[S]plit along with all associated training and evaluation datasets as open resources, enabling life scientists to immediately benefit from the potential of computational multiplexing and thus help accelerate the rate of their scientific discovery process.

bioinformatics↗

bioRxiv2024

HYDRA: HYdrogel Dispensing with Robotic Automation for high throughput drug testing

Traditional plastic- and glass-based cell culture platforms lack physiological relevance which limits their predictive power in drug development. Organoids and organs-on-chip offer biomimicry in late preclinical stages but lack scalability for high-throughput screening (HTS) in earlier stages. In fact, even introducing the simplest critical biomimetic cue, thin hydrogel layers, into the small wells of HTS plates causes the formation of curved menisci that disrupt cell seeding and imaging. We introduce HYDRA (HYDrogels by Robotic liquid handling Automation), an automated method that enables consistent hydrogel fabrication in HTS plates. We used liquid handling robots to dispense hydrogel precursor solutions and avoid contact with the sidewalls to prevent meniscus formation. We used gelatin for its mild surfactant properties, which allow pinning of the droplets contact line, promoting uniform spread. Finally, we re-aspirated the excess liquid, leaving behind a hydrogel layer with controlled stiffness and thickness. We show compatibility with standard 96- and 384-well plates and integration with existing automated workflows, including quality control. We validated HYDRA in dose-response assays using anti-cancer drugs, genetically engineered epithelial cells, phase holography, and fluorescence microscopy. We believe that HYDRA provides a scalable, biomimetic, and therefore more predictive alternative for early- stage screening and in vitro drug testing.

bioengineering↗

bioRxiv2024

CALIPERS: Cell cycle-aware live imaging for phenotyping experiments and regeneration studies

Cell cycle progression, migration, and proliferation shape development and regeneration, but simultaneous live-cell imaging remains challenging as conventional fluorescent cell cycle indicators (FUCCI) monopolize the green and red channels used by most structural and functional biosensors. To overcome this, we integrated a spectrally re-engineered FUCCI variant, open-source analysis software, and four-color human stem cell reporter lines into CALIPERS: a method for Cell-cycle-Aware Live-cell Imaging in Phenotyping and Regeneration Studies.

bioengineering↗