Substituting density functional theory in reaction barrier calculations for hydrogen atom transfer in proteins.

Hydrogen atom transfer (HAT) reactions are important in many biological systems. As these reactions are hard to observe experimentally, it is of high interest to shed light on them using simulations. Here, we present a machine learning model based on graph neural networks for the prediction of energy barriers of HAT reactions in proteins. As input, the model uses exclusively non-optimized structures as obtained from classical simulations. It was trained on more than 17 000 energy barriers calculated using hybrid density functional theory. We built and evaluated the model in the context of HAT in collagen, but we show that the same workflow can easily be applied to HAT reactions in other biological or synthetic polymers. We obtain for relevant reactions (small reaction distances) a model with good predictive power (R2 ∼ 0.9 and mean absolute error of <3 kcal mol-1). As the inference speed is high, this model enables evaluations of dozens of chemical situations within seconds. When combined with molecular dynamics in a kinetic Monte-Carlo scheme, the model paves the way toward reactive simulations.

Cell-free biosynthesis combined with deep learning accelerates de novo-development of antimicrobial peptides.

Bioactive peptides are key molecules in health and medicine. Deep learning holds a big promise for the discovery and design of bioactive peptides. Yet, suitable experimental approaches are required to validate candidates in high throughput and at low cost. Here, we established a cell-free protein synthesis (CFPS) pipeline for the rapid and inexpensive production of antimicrobial peptides (AMPs) directly from DNA templates. To validate our platform, we used deep learning to design thousands of AMPs de novo. Using computational methods, we prioritized 500 candidates that we produced and screened with our CFPS pipeline. We identified 30 functional AMPs, which we characterized further through molecular dynamics simulations, antimicrobial activity and toxicity. Notably, six de novo-AMPs feature broad-spectrum activity against multidrug-resistant pathogens and do not develop bacterial resistance. Our work demonstrates the potential of CFPS for high throughput and low-cost production and testing of bioactive peptides within less than 24 h.

A versatile active learning workflow for optimization of genetic and metabolic networks.

Optimization of biological networks is often limited by wet lab labor and cost, and the lack of convenient computational tools. Here, we describe METIS, a versatile active machine learning workflow with a simple online interface for the data-driven optimization of biological targets with minimal experiments. We demonstrate our workflow for various applications, including cell-free transcription and translation, genetic circuits, and a 27-variable synthetic CO2-fixation cycle (CETCH cycle), improving these systems between one and two orders of magnitude. For the CETCH cycle, we explore 1025 conditions with only 1,000 experiments to yield the most efficient CO2-fixation cascade described to date. Beyond optimization, our workflow also quantifies the relative importance of individual factors to the performance of a system identifying unknown interactions and bottlenecks. Overall, our workflow opens the way for convenient optimization and prototyping of genetic and metabolic networks with customizable adjustments according to user experience, experimental setup, and laboratory facilities.

Division and Regrowth of Phase-Separated Giant Unilamellar Vesicles*.

Success in the bottom-up assembly of synthetic cells will depend on strategies for the division of protocellular compartments. Here, we describe the controlled division of phase-separated giant unilamellar lipid vesicles (GUVs). We derive an analytical model based on the vesicle geometry, which makes four quantitative predictions that we verify experimentally. We find that the osmolarity ratio required for division is 2 , independent of the GUV size, while asymmetric division happens at lower osmolarity ratios. Remarkably, we show that a suitable osmolarity change can be triggered by water evaporation, enzymatic decomposition of sucrose or light-triggered uncaging of CMNB-fluorescein. The latter provides full spatiotemporal control, such that a target GUV undergoes division whereas the surrounding GUVs remain unaffected. Finally, we grow phase-separated vesicles from single-phased vesicles by targeted fusion of the opposite lipid type with programmable DNA tags to enable subsequent division cycles.

Recruitment of 53BP1 Proteins for DNA Repair and Persistence of Repair Clusters Differ for Cell Types as Detected by Single Molecule Localization Microscopy.

DNA double stranded breaks (DSBs) are the most serious type of lesions introduced into chromatin by ionizing radiation. During DSB repair, cells recruit different proteins to the damaged sites in a manner dependent on local chromatin structure, DSB location in the nucleus, and the repair pathway entered. 53BP1 is one of the important players participating in repair pathway decision of the cell. Although many molecular biology details have been investigated, the architecture of 53BP1 repair foci and its development during the post-irradiation time, especially the period of protein recruitment, remains to be elucidated. Super-resolution light microscopy is a powerful new tool to approach such studies in 3D-conserved cell nuclei. Recently, we demonstrated the applicability of single molecule localization microscopy (SMLM) as one of these highly resolving methods for analyses of dynamic repair protein distribution and repair focus internal nano-architecture in intact cell nuclei. In the present study, we focused our investigation on 53BP1 foci in differently radio-resistant cell types, moderately radio-resistant neonatal human dermal fibroblasts (NHDF) and highly radio-resistant U87 glioblastoma cells, exposed to high-LET 15N-ion radiation. At given time points up to 24 h post irradiation with doses of 1.3 Gy and 4.0 Gy, the coordinates and spatial distribution of fluorescently tagged 53BP1 molecules was quantitatively evaluated at the resolution of 10⁻20 nm. Clusters of these tags were determined as sub-units of repair foci according to SMLM parameters. The formation and relaxation of such clusters was studied. The higher dose generated sufficient numbers of DNA breaks to compare the post-irradiation dynamics of 53BP1 during DSB processing for the cell types studied. A perpendicular (90°) irradiation scheme was used with the 4.0 Gy dose to achieve better separation of a relatively high number of particle tracks typically crossing each nucleus. For analyses along ion-tracks, the dose was reduced to 1.3 Gy and applied in combination with a sharp angle irradiation (10° relative to the cell plane). The results reveal a higher ratio of 53BP1 proteins recruited into SMLM defined clusters in fibroblasts as compared to U87 cells. Moreover, the speed of foci and thus cluster formation and relaxation also differed for the cell types. In both NHDF and U87 cells, a certain number of the detected and functionally relevant clusters remained persistent even 24 h post irradiation; however, the number of these clusters again varied for the cell types. Altogether, our findings indicate that repair cluster formation as determined by SMLM and the relaxation (i.e., the remaining 53BP1 tags no longer fulfill the cluster definition) is cell type dependent and may be functionally explained and correlated to cell specific radio-sensitivity. The present study demonstrates that SMLM is a highly appropriate method for investigations of spatiotemporal protein organization in cell nuclei and how it influences the cell decision for a particular repair pathway at a given DSB site.