Comparison of High- and Low-LET Radiation-Induced DNA Double-Strand Break Processing in Living Cells.

High-linear-energy-transfer (LET) radiation is more lethal than similar doses of low-LET radiation types, probably a result of the condensed energy deposition pattern of high-LET radiation. Here, we compare high-LET α-particle to low-LET X-ray irradiation and monitor double-strand break (DSB) processing. Live-cell microscopy was used to monitor DNA double-strand breaks (DSBs), marked by p53-binding protein 1 (53BP1). In addition, the accumulation of the endogenous 53BP1 and replication protein A (RPA) DSB processing proteins was analyzed by immunofluorescence. In contrast to α-particle-induced 53BP1 foci, X-ray-induced foci were resolved quickly and more dynamically as they showed an increase in 53BP1 protein accumulation and size. In addition, the number of individual 53BP1 and RPA foci was higher after X-ray irradiation, while focus intensity was higher after α-particle irradiation. Interestingly, 53BP1 foci induced by α-particles contained multiple RPA foci, suggesting multiple individual resection events, which was not observed after X-ray irradiation. We conclude that high-LET α-particles cause closely interspaced DSBs leading to high local concentrations of repair proteins. Our results point toward a change in DNA damage processing toward DNA end-resection and homologous recombination, possibly due to the depletion of soluble protein in the nucleoplasm. The combination of closely interspaced DSBs and perturbed DNA damage processing could be an explanation for the increased relative biological effectiveness (RBE) of high-LET α-particles compared to X-ray irradiation.

The Power of Universal Contextualized Protein Embeddings in Cross-species Protein Function Prediction.

Computationally annotating proteins with a molecular function is a difficult problem that is made even harder due to the limited amount of available labeled protein training data. Unsupervised protein embeddings partly circumvent this limitation by learning a universal protein representation from many unlabeled sequences. Such embeddings incorporate contextual information of amino acids, thereby modeling the underlying principles of protein sequences insensitive to the context of species. We used an existing pre-trained protein embedding method and subjected its molecular function prediction performance to detailed characterization, first to advance the understanding of protein language models, and second to determine areas of improvement. Then, we applied the model in a transfer learning task by training a function predictor based on the embeddings of annotated protein sequences of one training species and making predictions on the proteins of several test species with varying evolutionary distance. We show that this approach successfully generalizes knowledge about protein function from one eukaryotic species to various other species, outperforming both an alignment-based and a supervised-learning-based baseline. This implies that such a method could be effective for molecular function prediction in inadequately annotated species from understudied taxonomic kingdoms.