Single-cell RNA sequencing (scRNA-seq) has become an essential tool for characterizing gene expression in eukaryotes, but current methods are incompatible with bacteria. Here, we introduce microSPLiT (microbial split-pool ligation transcriptomics), a high-throughput scRNA-seq method for Gram-negative and Gram-positive bacteria that can resolve heterogeneous transcriptional states. We applied microSPLiT to >25,000 Bacillus subtilis cells sampled at different growth stages, creating an atlas of changes in metabolism and lifestyle. We retrieved detailed gene expression profiles associated with known, but rare, states such as competence and prophage induction and also identified unexpected gene expression states, including the heterogeneous activation of a niche metabolic pathway in a subpopulation of cells. MicroSPLiT paves the way to high-throughput analysis of gene expression in bacterial communities that are otherwise not amenable to single-cell analysis, such as natural microbiota.

The design of modular protein logic for regulating protein function at the posttranscriptional level is a challenge for synthetic biology. Here, we describe the design of two-input AND, OR, NAND, NOR, XNOR, and NOT gates built from de novo–designed proteins. These gates regulate the association of arbitrary protein units ranging from split enzymes to transcriptional machinery in vitro, in yeast and in primary human T cells, where they control the expression of the TIM3 gene related to T cell exhaustion. Designed binding interaction cooperativity, confirmed by native mass spectrometry, makes the gates largely insensitive to stoichiometric imbalances in the inputs, and the modularity of the approach enables ready extension to three-input OR, AND, and disjunctive normal form gates. The modularity and cooperativity of the control elements, coupled with the ability to de novo design an essentially unlimited number of protein components, should enable the design of sophisticated posttranslational control logic over a wide range of biological functions.

Deep learning generative approaches provide an opportunity to broadly explore protein structure space beyond the sequences and structures of natural proteins. Here we use deep network hallucination to generate a wide range of symmetric protein homo-oligomers given only a specification of the number of protomers and the protomer length. Crystal structures of 7 designs are very close to the computational models (median RMSD: 0.6 Å), as are 3 cryoEM structures of giant 10 nanometer rings with up to 1550 residues and C33 symmetry; all differ considerably from previously solved structures. Our results highlight the rich diversity of new protein structures that can be generated using deep learning, and pave the way for the design of increasingly complex components for nanomachines and biomaterials.

Summary Ruminococcus bromii is a dominant member of the human colonic microbiota that plays a ‘keystone’ role in degrading dietary resistant starch. Recent evidence from one strain has uncovered a unique cell surface ‘amylosome’ complex that organizes starch‐degrading enzymes. New genome analysis presented here reveals further features of this complex and shows remarkable conservation of amylosome components between human colonic strains from three different continents and a R. bromii strain from the rumen of Australian cattle. These R. bromii strains encode a narrow spectrum of carbohydrate active enzymes (CAZymes) that reflect extreme specialization in starch utilization. Starch hydrolysis products are taken up mainly as oligosaccharides, with only one strain able to grow on glucose. The human strains, but not the rumen strain, also possess transporters that allow growth on galactose and fructose. R. bromii strains possess a full complement of sporulation and spore germination genes and we demonstrate the ability to form spores that survive exposure to air. Spore formation is likely to be a critical factor in the ecology of this nutritionally highly specialized bacterium, which was previously regarded as ‘non‐sporing’, helping to explain its widespread occurrence in the gut microbiota through the ability to transmit between hosts.

Single-cell RNA-sequencing (scRNA-seq) has become an essential tool for characterizing multicelled eukaryotic systems but current methods are not compatible with bacteria. Here, we introduce microSPLiT, a low cost and high-throughput scRNA-seq method that works for gramnegative and gram-positive bacteria and can resolve transcriptional states that remain hidden at a population level. We applied microSPLiT to >25,000 Bacillus subtilis cells sampled from different growth stages, creating a detailed atlas of changes in metabolism and lifestyle. We not only retrieve detailed gene expression profiles associated with known but rare states such as competence and PBSX prophage induction, but also identify novel and unexpected gene expression states including heterogeneous activation of a niche metabolic pathway in a subpopulation of cells. microSPLiT empowers high-throughput analysis of gene expression in complex bacterial communities.

No abstract

Reverse osmosis (RO) is a membrane technology that separates dissolved species from water. RO has been applied for the removal of chemical contaminants from water for potable reuse applications. The presence of a wide variety of influent chemical contaminants and the insufficient rejection of low-molecular-weight neutral organics by RO calls for the need to develop a model that predicts the rejection of various organics. In this study, we develop a group contribution method (GCM) to predict the mass transfer coefficients by fragmenting the structure of low-molecular-weight neutral organics into small parts that interact with the RO membrane. Overall, 54 organics including 26 halogenated and oxygenated alkanes, 8 alkenes, and 20 alkyl and halobenzenes were used to determine 39 parameters to calibrate for 6 different RO membranes, including 4 brackish water and 2 seawater membranes. Through six membranes, approximately 80% of calculated rejection was within an error goal (i.e., ±5%) from the experimental observation. To extend the GCM for a reference RO membrane, ESPA2-LD, 14 additional organics were included from the literature to calibrate nitrogen-containing functional groups of nitrosamine, nitriles, and amide compounds. Overall, 49 organics (72% of 68 compounds) from calibration and 7 compounds (87.5% of 8 compounds) from prediction were within the error goal.

Deep learning generative approaches provide an opportunity to broadly explore protein structure space beyond the sequences and structures of natural proteins. Here we use deep network hallucination to generate a wide range of symmetric protein homo-oligomers given only a specification of the number of protomers and the protomer length. Crystal structures of 7 designs are very close to the computational models (median RMSD: 0.6 Å), as are 3 cryoEM structures of giant rings with up to 1550 residues, C33 symmetry, and 10 nanometer in diameter; all differ considerably from previously solved structures. Our results highlight the rich diversity of new protein structures that can be created using deep learning, and pave the way for the design of increasingly complex nanomachines and biomaterials.