Biology subjects

Kratz, J.

Publications and source records attributed to Kratz, J..

3 recordsLinked to original sources

bioRxiv2026

Epigenetic de-repression of basal cell metaplasia in aging AT2 cells is a risk factor for idiopathic pulmonary fibrosis (IPF).

Idiopathic pulmonary fibrosis (IPF) is a fatal, age-associated lung disease in which alveolar type II (AT2) cells lose regenerative capacity and can adopt aberrant basal-like fates that promote fibrosis. Using 3D organoid co-cultures with primary human fibroblasts, we find that healthy human AT2 cell trans-differentiation into KRT5+/KRT17+ basal cells increases progressively with age, while differentiation into RAGE+ AT1-like cells decreases. We identify a shared gene signature in AT2 cells at downstream targets of p63 characterized both by acquisition of bivalent, poised chromatin marks with age and increased accessibility in IPF, indicating epigenetic "priming" towards a basal cell lineage. In vitro treatment of young AT2 cells with IL-1{beta} recapitulates this priming toward basal differentiation via a NF-kB-regulated histone demethylase, JMJD3. Conversion of primed AT2 cells to a basal fate requires recruitment of a shared transcription factor, KLF5, from AT1-specific to basal-specific promoters by HIF-1. AT2 cells instead convert to KRT5-/KRT17+ basaloid cells via a non-age-dependent pathway that requires KLF5-SMAD2/3 complexing through TGF{beta}1 signaling. These findings define an inflammation-driven epigenetic de-repressive mechanism that links aging, inflammatory stress, hypoxia, and dysfunctional epithelial metaplasia, and accounts for the likely origin of aberrant epithelial cell populations in fibrotic lung disease.

cell biology↗

bioRxiv2025

Reinforcement learning for adaptive control of phenotypically heterogeneous bacterial populations

Bacterial populations display extraordinary resilience to antibiotic stress, driven by diverse physiological states that allow some cells to persist and later repopulate. This phenotypic heterogeneity, amplified by environmental fluctuations, undermines the effectiveness of conventional fixed-dose treatment regimens. To address this challenge, we introduce a reinforcement learning (RL) framework that discovers adaptive treatment strategies using only experimentally accessible, population-level measurements. The RL agent learns to infer the hidden physiological state of the population and leverages this knowledge to maintain control even under conditions not encountered during training. Moreover, when granted control over nutrient availability, an important driver of physiological change often overlooked in antibiotic treatment protocols, the agent consistently drives population extinction, surpassing adaptive protocols based solely on drug dynamics. This computational framework offers a powerful, data-driven approach for designing adaptive treatment strategies to counter the growing threat of antimicrobial resistance.

bioinformatics↗

bioRxiv2025

Power-law memory governs bacterial adaptation and learning in fluctuating environments

How do single-celled organisms adapt and learn to survive in dynamic environments without a nervous system? Here, we provide experimental evidence and a theoretical model demon-strating learning-like behavior by single bacterial cells in fluctuating environments. Using a custom microfluidic platform, we tracked individual E. coli cells in dynamic nutrient environments and found that bacteria adapt on multiple timescales, tuning their growth control behavior based on prior environmental experience. Motivated by our observation that cellular adaptation dynamics are approximately scale-free, we built a theoretical framework for bacterial growth control with dynamic power-law memory to explain how bacteria integrate environmental information over a range of timescales to enable growth rate adaptation. We show how this behavior arises naturally from heterogeneous ribosomal relaxation dynamics within a bacterial cell. Using this model, we identify an inherent trade-off between growth rate maximization and adaptation speed, which we validate experimentally in pulsatile nutrient environments. Finally, we connect our mechanistic reaction-network model to descriptions of artificial recurrent neural networks, identifying a minimal network architecture capable of exhibiting adaptation and learning at the single-cell level.

systems biology↗