Label-free ultraviolet photoacoustic microscopy for multiscale structural phenotyping of epithelial morphogenesis in organoid-derived intestinal epithelial wrinkle tissue.

Mechanical wrinkling of epithelial tissues encodes key features of tissue morphogenesis and mechanotransductive responses, yet its quantitative, multiscale characterization remains challenging in organoid-based in vitro models due to reliance on fluorescence labelling and sectioning. Here, we develop label-free ultraviolet photoacoustic microscopy (UV-PAM) for multiscale imaging and quantitative analysis of human organoid-derived epithelial wrinkle tissues. By exploiting intrinsic nucleic-acid absorption at 266 nm, UV-PAM visualizes wrinkle structures and nuclear morphology without labeling or sectioning. Using morphomechanically defined wrinkle tissues, we identify multiscale signatures, including wrinkle wavelength, wrinkle index, and strain-induced nuclear elongation and redistribution. UV-PAM further captures chemically induced epithelial stress, revealing degradation of wrinkle structure and nuclear morphology. In a proof-of-concept dataset comprising two untreated and two DTT-treated specimens, integration of these intrinsic signatures with deep learning enabled patch-level discrimination between untreated and chemically perturbed wrinkle tissues on held-out field of views (FOVs), with training and test FOVs obtained from the same specimens. These results establish UV-PAM as a label-free platform for multiscale imaging of epithelial morphogenesis and preliminary computational phenotyping.