Biology subjects

Han, K. S.

Publications and source records attributed to Han, K. S..

5 recordsLinked to original sources

bioRxiv2026

Virtual multiplex staining of the pancreatic islets across type 1 diabetes progression using a Schroedinger bridge

Classical hematoxylin and eosin (H&E) staining enables review of tissue morphology but lacks information regarding the molecular state of cells. Immunohistochemical (IHC) techniques label specific proteins in tissue, allowing differentiation of relevant structures that may go undetectable in H&E. However, the IHC process is complex, expensive, and time-consuming, especially for multiplex IHC (mIHC) limiting its use in large cohorts. Stain conversion of H&E to IHC using generative artificial intelligence models such as generative adversarial networks (GANs) represent one solution to this problem. However, GANs are unstable during out of distribution sampling and are prone to hallucinations or mode collapse, limiting their accuracy in challenging image conversion tasks. To address this, the field has recently turned to diffusion models. Here, we introduce Schrodinger-bridge for Multiplex ImmunoLabel Estimation (SMILE). Unlike conventional diffusion models that map from source to target through an intermediate Gaussian noise, Schrodinger-bridge diffusion models skip this step and have been shown to better preserve structures during image translation. To test the performance of SMILE, we generated a large cohort of high-fidelity H&E-mIHC image pairs from pancreatic organ donors, targeting insulin, glucagon, and CD3. Our dataset well-sampled across type-1 diabetes status, pancreas anatomical location, age, and sex. Using this cohort, we demonstrate the superiority of SMILE compared to GANs via a comprehensive evaluation framework incorporating texture, distribution, and antibody-specific metrics, as well as blinded pathologist reviews. We further confirmed the ability of SMILE to generate accurate mIHC images from H&Es generated at an external site, to perform whole slide image conversion, and to generate realistic three-dimensional maps of the pancreatic islets in non-diabetic, auto-antibody positive, and type-1 diabetic donor tissue. Finally, we performed stain conversion of paired H&E to HER2 and Ki67 images in breast cancer, confirming the superiority of SMILE in diverse stain conversion applications. Collectively, this framework provides a scalable pipeline for high-throughput proteomic inference from archival H&Es, providing transformative potential for pancreatic research and digital pathology.

bioinformatics↗

bioRxiv2025

Integration of nuclear morphology and 3D imaging to profile cellular neighborhoods

Nuclear morphology is an indicator of cellular function and disease states, as changes in nuclear size, shape, and texture often reflect underlying disease-related genetic, epigenetic, and microenvironmental alterations. For disease diagnosis, nuclear segmentation performed in 2D hematoxylin and eosin (H&E)-stained tissue sections has long represented the gold standard. However, recent advances in three-dimensional (3D) histology, which provide a more biologically accurate representation of the spatial heterogeneity of human microanatomy, has led to improved understandings of disease pathology. Yet challenges remain in the development of scalable and computationally efficient pipelines for extracting and interpreting nuclear features in 3D space. 2D histology neglects crucial spatial information, such as 3D connectivity, morphology, and rare events missed by sparser sampling. Here, through extension of the CODA platform, we integrate 3D imaging with nuclear segmentation to analyze nuclear morphological features in human tissue. Analysis of 3D tissue microenvironments uncovered critical changes in 3D morphometric heterogeneity. Additionally, it enables the spatial characterization of immune cell distribution in relation to tissue structures, such as variations in leukocyte density near pancreatic ducts and blood vessels of different sizes. This approach provides a more comprehensive understanding of tissue and nuclear structures, revealing spatial patterns and interactions that are critical for disease progression.

bioinformatics↗

bioRxiv2024

The microanatomy of human skin in aging

Aging is a major driver of diseases in humans. Identifying features associated with aging is essential for designing robust intervention strategies and discovering novel biomarkers of aging. Extensive studies at both the molecular and organ/whole-body physiological scales have helped determined features associated with aging. However, the lack of meso-scale studies, particularly at the tissue level, limits the ability to translate findings made at molecular scale to impaired tissue functions associated with aging. In this work, we established a tissue image analysis workflow - quantitative micro-anatomical phenotyping (qMAP) - that leverages deep learning and machine vision to fully label tissue and cellular compartments in tissue sections. The fully mapped tissue images address the challenges of finding an interpretable feature set to quantitatively profile age-related microanatomic changes. We optimized qMAP for skin tissues and applied it to a cohort of 99 donors aged 14 to 92. We extracted 914 microanatomic features and found that a broad spectrum of these features, represented by 10 cores processes, are strongly associated with aging. Our analysis shows that microanatomical features of the skin can predict aging with a mean absolute error (MAE) of 7.7 years, comparable to state-of-the-art epigenetic clocks. Our study demonstrates that tissue-level architectural changes are strongly associated with aging and represent a novel category of aging biomarkers that complement molecular markers. Our results highlight the complex and underexplored multi-scale relationship between molecular and tissue microanatomic scales.

pathology↗

bioRxiv2024

Substrate stiffness modulates the emergence and magnitude of senescence phenotypes

Cellular senescence is a major driver of aging and disease. Here we show that substrate stiffness modulates the emergence and magnitude of senescence phenotypes after exposure to senescence inducers. Using a primary dermal fibroblast model, we show that decreased substrate stiffness accelerates senescence-associated cell-cycle arrest and regulates the expression of conventional protein-based biomarkers of senescence. We found that the expression of these senescence biomarkers, namely p21WAF1/CIP1 and p16INK4a are mechanosensitive and are in-part regulated by myosin contractility through focal adhesion kinase (FAK)-ROCK signaling. Interestingly, at the protein level senescence-induced dermal fibroblasts on soft substrates (0.5 kPa) do not express p21WAF1/CIP1 and p16INK4a at comparable levels to induced cells on stiff substrates (4GPa). However, cells express CDKN1a, CDKN2a, and IL6 at the RNA level across both stiff and soft substrates. Moreover, when cells are transferred from soft to stiff substrates, senescent cells recover an elevated expression of p21WAF1/CIP1 and p16INK4a at levels comparable to senescence cells on stiff substrates, pointing to a mechanosensitive regulation of the senescence phenotype. Together, our results indicate that the emergent senescence phenotype depends critically on the local mechanical environments of cells and that senescent cells actively respond to changing mechanical cues.

molecular biology↗

bioRxiv2020

In situ characterization of the 3D microanatomy of the pancreas and pancreatic cancer at single cell resolution

Pancreatic ductal adenocarcinoma (PDAC) is one of the deadliest forms of cancer. Accumulating evidence indicates the tumor microenvironment is highly associated with tumorigenesis through regulation of cellular physiology, signaling systems, and gene expression profiles of cancer cells. Yet the mechanisms by which the microenvironment evolves from normal pancreas architecture to precursor lesions and invasive cancer is poorly understood. Obtaining high-content and high-resolution information from a complex tumor microenvironment in large volumetric landscapes represents a key challenge in the field of cancer biology. To address this challenge, we established a novel method to reconstruct three-dimensional (3D) centimeter-scale tissues containing billions of cells from serially sectioned histological samples, utilizing deep learning approaches to recognize eight distinct tissue subtypes from hematoxylin and eosin stained sections at micrometer and single-cell resolution. Using samples from a range of normal, precancerous, and invasive pancreatic cancer tissue, we map in 3D modes of cancer invasion in the tumor microenvironment, and emphasize the need for further 3D quantification of biological systems.

cancer biology↗