Freund, D. | Search bioRxiv

bioRxiv2025

A synthetic bacterium that degrades and assimilates poly(ethylene terephthalate)

Polyethylene terephthalate (PET) is the fourth most commonly used plastic worldwide. Like all plastics, post-consumer PET is poorly managed and accumulates in the environment, posing significant ecological threats. After 70 years of accumulation, microorganisms capable of degrading and assimilating PET have been isolated, demonstrating that PET can be broken down and converted into valuable cellular biomass or metabolic products. These natural isolates, however, are poorly characterized and challenging to genetically manipulate, which limits their further optimization and applicability. Here, we engineer a well-established synthetic biology chassis for the biodegradation and assimilation of PET. We modified the bacterium Pseudomonas putida KT2440 to heterologously express an active PET-hydrolytic enzyme extracellularly and to metabolize PET biodegradation products. The resulting strain, named PETBuster, was capable of growing on PET as the sole carbon source on solid and liquid media. We achieved 91% PET degradation after 21 days of culture, with a doubling time of 3.6 days, under mesophilic conditions. In this way, we demonstrate that PET fermentation is feasible, opening the door to the production of valuable chemicals from waste.

synthetic biology↗

bioRxiv2025

Zero-Shot, Big-Shot, Active-Shot - How to estimate cell confluence, lazily

Mesenchymal stem cell therapy shows promising results for difficult-to-treat diseases, but standardized manufacturing requires robust quality control through automated cell confluence monitoring. While deep learning can automate confluence estimation, research on cost-effective dataset curation and the role of foundation models in this task remains limited. We systematically investigate the most effective strategies for confluence estimation, focusing on active learning-based dataset curation, goal-specific labeling, and leveraging foundation models for zero-shot inference. Here, we show that zero-shot inference with the Segment Anything Model (SAM) achieves excellent confluence estimation without any task-specific training, outperforming fine-tuned smaller models. Further, our findings demonstrate that active learning does not significantly improve model dataset curation compared to random selection in homogeneous cell datasets. We show that goal-specific, simplified labeling strategies perform comparably to precise annotations while substantially reducing annotation effort. These results challenge common assumptions about dataset curation: neither active learning nor extensive fine-tuning provided significant benefits for our specific use case. Instead, we found that leveraging SAMs zero-shot capabilities and targeted labeling strategies offers the most cost-effective approach to automated confluence estimation. Our work provides practical guidelines for implementing automated cell monitoring in MSC manufacturing, demonstrating that extensive dataset curation may be unnecessary when foundation models can effectively handle the task out of the box.

bioinformatics↗