Hidden protein functions and what they may teach us

Machine learning-aided design and screening of an emergent protein function in synthetic cells

Recently, utilization of Machine Learning (ML) has led to astonishing progress in computational protein design, bringing into reach the targeted engineering of proteins for industrial and biomedical applications. However, the design of proteins for emergent functions of core relevance to cells, such as the ability to spatiotemporally self-organize and thereby structure the cellular space, is still extremely challenging. While on the generative side conditional generative models and multi-state design are on the rise, for emergent functions there is a lack of tailored screening methods as typically needed in a protein design project, both computational and experimental. Here we describe a proof-of-principle of how such screening, in silico and in vitro, can be achieved for ML-generated variants of a protein that forms intracellular spatiotemporal patterns. For computational screening we use a structure-based divide-and-conquer approach to find the most promising candidates, while for the subsequent in vitro screening we use synthetic cell-mimics as established by Bottom-Up Synthetic Biology. We then show that the best screened candidate can indeed completely substitute the wildtype gene in Escherichia coli. These results raise great hopes for the next level of synthetic biology, where ML-designed synthetic proteins will be used to engineer cellular functions.

Perspectives on polarity - exploring biological asymmetry across scales

In this Perspective, Journal of Cell Science invited researchers working on cell and tissue polarity to share their thoughts on unique, emerging or open questions relating to their field. The goal of this article is to feature 'voices' from scientists around the world and at various career stages, to bring attention to innovative and thought-provoking topics of interest to the cell biology community. These voices discuss intriguing questions that consider polarity across scales, evolution, development and disease. What can yeast and protists tell us about the evolution of cell and tissue polarity in animals? How are cell fate and development influenced by emerging dynamics in cell polarity? What can we learn from atypical and extreme polarity systems? How can we arrive at a more unified biophysical understanding of polarity? Taken together, these pieces demonstrate the broad relevance of the fascinating phenomenon of cell polarization to diverse fundamental biological questions.

Protein-Based Patterning to Spatially Functionalize Biomimetic Membranes

The bottom-up reconstitution of proteins for their modular engineering into synthetic cellular systems can reveal hidden protein functions in vitro. This is particularly evident for the bacterial Min proteins, a paradigm for self-organizing reaction-diffusion systems that displays an unexpected functionality of potential interest for bioengineering: the directional active transport of any diffusible cargo molecule on membranes. Here, the MinDE protein system is reported as a versatile surface patterning tool for the rational design of synthetically assembled 3D systems. Employing two-photon lithography, microswimmer-like structures coated with tailored lipid bilayers are fabricated and demonstrate that Min proteins can uniformly pattern bioactive molecules on their surface. Moreover, it is shown that the MinDE system can form stationary patterns inside lipid vesicles, which allow the targeting and distinctive clustering of higher-order protein structures on their inner leaflet. Given their facile use and robust function, Min proteins thus constitute a valuable molecular toolkit for spatially patterned functionalization of artificial biosystems like cell mimics and microcarriers.

The Intransitive Logic of Directed Cycles and Flipons Enhances the Evolution of Molecular Computers by Augmenting the Kolmogorov Complexity of Genomes

Cell responses are usually viewed as transitive events with fixed inputs and outputs that are regulated by feedback loops. In contrast, directed cycles (DCs) have all nodes connected, and the flow is in a single direction. Consequently, DCs can regenerate themselves and implement intransitive logic. DCs are able to couple unrelated chemical reactions to each edge. The output depends upon which node is used as input. DCs can also undergo selection to minimize the loss of thermodynamic entropy while maximizing the gain of information entropy. The intransitive logic underlying DCs enhances their programmability and impacts their evolution. The natural selection of DCs favors the persistence, adaptability, and self-awareness of living organisms and does not depend solely on changes to coding sequences. Rather, the process can be RNA-directed. I use flipons, nucleic acid sequences that change conformation under physiological conditions, as a simple example and then describe more complex DCs. Flipons are often encoded by repeats and greatly increase the Kolmogorov complexity of genomes by adopting alternative structures. Other DCs allow cells to regenerate, recalibrate, reset, repair, and rewrite themselves, going far beyond the capabilities of current computational devices. Unlike Turing machines, cells are not designed to halt but rather to regenerate.

Toward the Complete Functional Characterization of a Minimal Bacterial Proteome

Recently, we presented a whole-cell kinetic model of the genetically minimal bacterium JCVI-syn3A that described the coupled metabolic and genetic information processes and predicted behaviors emerging from the interactions among these networks. JCVI-syn3A is a genetically reduced bacterial cell that has the fewest number and smallest fraction of genes of unclear function, with approximately 90 of its 452 protein-coding genes (that is less than 20%) unannotated. Further characterization of unclear JCVI-syn3A genes strengthens the robustness and predictive power of cell modeling efforts and can lead to a deeper understanding of biophysical processes and pathways at the cell scale. Here, we apply computational analyses to elucidate the functions of the products of several essential but previously uncharacterized genes involved in integral cellular processes, particularly those directly affecting cell growth, division, and morphology. We also suggest directed wet-lab experiments informed by our analyses to further understand these "missing puzzle pieces" that are an essential part of the mosaic of biological interactions present in JCVI-syn3A. Our workflow leverages evolutionary sequence analysis, protein structure prediction, interactomics, and genome architecture to determine upgraded annotations. Additionally, we apply the structure prediction analysis component of our work to all 452 protein coding genes in JCVI-syn3A to expedite future functional annotation studies as well as the inverse mapping of the cell state to more physical models requiring all-atom or coarse-grained representations for all JCVI-syn3A proteins.