Essential and virulence-related protein interactions of pathogens revealed through deep learning.

Identification of bacterial protein-protein interactions and predicting the structures of the complexes could aid in the understanding of pathogenicity mechanisms and developing treatments for infectious diseases. Here, we developed a deep learning-based pipeline that leverages residue-residue coevolution and protein structure prediction to systematically identify and structurally characterize protein-protein interactions at the proteome-wide scale. Using this pipeline, we searched through 78 million pairs of proteins across 19 human bacterial pathogens and identified 1923 confidently predicted complexes involving essential genes and 256 involving virulence factors. Many of these complexes were not previously known; we experimentally tested 12 such predictions, and half of them were validated. The predicted interactions span core metabolic and virulence pathways ranging from post-transcriptional modification to acid neutralization to outer membrane machinery and should contribute to our understanding of the biology of these important pathogens and the design of drugs to combat them.

De novo design of pH-responsive self-assembling helical protein filaments.

Biological evolution has led to precise and dynamic nanostructures that reconfigure in response to pH and other environmental conditions. However, designing micrometre-scale protein nanostructures that are environmentally responsive remains a challenge. Here we describe the de novo design of pH-responsive protein filaments built from subunits containing six or nine buried histidine residues that assemble into micrometre-scale, well-ordered fibres at neutral pH. The cryogenic electron microscopy structure of an optimized design is nearly identical to the computational design model for both the subunit internal geometry and the subunit packing into the fibre. Electron, fluorescent and atomic force microscopy characterization reveal a sharp and reversible transition from assembled to disassembled fibres over 0.3 pH units, and rapid fibre disassembly in less than 1 s following a drop in pH. The midpoint of the transition can be tuned by modulating buried histidine-containing hydrogen bond networks. Computational protein design thus provides a route to creating unbound nanomaterials that rapidly respond to small pH changes.

Atomically accurate de novo design of single-domain antibodies.

Despite the central role that antibodies play in modern medicine, there is currently no way to rationally design novel antibodies to bind a specific epitope on a target. Instead, antibody discovery currently involves time-consuming immunization of an animal or library screening approaches. Here we demonstrate that a fine-tuned RFdiffusion network is capable of designing de novo antibody variable heavy chains (VHH's) that bind user-specified epitopes. We experimentally confirm binders to four disease-relevant epitopes, and the cryo-EM structure of a designed VHH bound to influenza hemagglutinin is nearly identical to the design model both in the configuration of the CDR loops and the overall binding pose.

Life goes on: Spatial heterogeneity promotes biodiversity in an urbanized coastal marine ecosystem.

Both human populations and marine biodiversity are concentrated along coastlines, with growing conservation interest in how these ecosystems can survive intense anthropogenic impacts. Tropical urban centres provide valuable research opportunities because these megacities are often adjacent to mega-diverse coral reef systems. The Pearl River Delta is a prime exemplar, as it encompasses one of the most densely populated and impacted regions in the world and is located just northwest of the Coral Triangle. However, the spatial and taxonomic complexity of this biodiversity, most of which is small, cryptic in habitat and poorly known, make comparative analyses challenging. We deployed standardized settlement structures at seven sites differing in the intensity of human impacts and used COI metabarcoding to characterize benthic biodiversity, with a focus on metazoans. We found a total of 7184 OTUs, with an average of 665 OTUs per sampling unit; these numbers exceed those observed in many previous studies using comparable methods, despite the location of our study in an urbanized environment. Beta diversity was also high, with 52% of the OTUs found at just one site. As expected, we found that the sites close to point sources of pollution had substantially lower diversity (44% less) relative to sites bathed in less polluted oceanic waters. However, the polluted sites contributed substantially to the total animal diversity of the region, with 25% of all OTUs occurring only within polluted sites. Further analysis of Arthropoda, Annelida and Mollusca showed that phylogenetic clustering within a site was common, suggesting that environmental filtering reduced biodiversity to a subset of lineages present within the region, a pattern that was most pronounced in polluted sites and for the Arthropoda. The water quality gradients surrounding the PRD highlight the unique role of in situ studies for understanding the impacts of complex urbanization pressures on biodiversity.

Hydrophobic mismatch drives self-organization of designer proteins into synthetic membranes.

The organization of membrane proteins between and within membrane-bound compartments is critical to cellular function. Yet we lack approaches to regulate this organization in a range of membrane-based materials, such as engineered cells, exosomes, and liposomes. Uncovering and leveraging biophysical drivers of membrane protein organization to design membrane systems could greatly enhance the functionality of these materials. Towards this goal, we use de novo protein design, molecular dynamic simulations, and cell-free systems to explore how membrane-protein hydrophobic mismatch could be used to tune protein cotranslational integration and organization in synthetic lipid membranes. We find that membranes must deform to accommodate membrane-protein hydrophobic mismatch, which reduces the expression and co-translational insertion of membrane proteins into synthetic membranes. We use this principle to sort proteins both between and within membranes, thereby achieving one-pot assembly of vesicles with distinct functions and controlled split-protein assembly, respectively. Our results shed light on protein organization in biological membranes and provide a framework to design self-organizing membrane-based materials with applications such as artificial cells, biosensors, and therapeutic nanoparticles.

Stimulating intestinal GIP release reduces food intake and body weight in mice.

OBJECTIVE: Glucose dependent insulinotropic polypeptide (GIP) is well established as an incretin hormone, boosting glucose-dependent insulin secretion. However, whilst anorectic actions of its sister-incretin glucagon-like peptide-1 (GLP-1) are well established, a physiological role for GIP in appetite regulation is controversial, despite the superior weight loss seen in preclinical models and humans with GLP-1/GIP dual receptor agonists compared with GLP-1R agonism alone. METHODS: We generated a mouse model in which GIP expressing K-cells can be activated through hM3Dq Designer Receptor Activated by Designer Drugs (DREADD, GIP-Dq) to explore physiological actions of intestinally-released GIP. RESULTS: In lean mice, Dq-stimulation of GIP expressing cells increased plasma GIP to levels similar to those found postprandially. The increase in GIP was associated with improved glucose tolerance, as expected, but also triggered an unexpected robust inhibition of food intake. Validating that this represented a response to intestinally-released GIP, the suppression of food intake was prevented by injecting mice peripherally or centrally with antagonistic GIPR-antibodies, and was reproduced in an intersectional model utilising Gip-Cre / Villin-Flp to limit Dq transgene expression to K-cells in the intestinal epithelium. The effects of GIP cell activation were maintained in diet induced obese mice, in which chronic K-cell activation reduced food intake and attenuated body weight gain. CONCLUSIONS: These studies establish a physiological gut-brain GIP-axis regulating food intake in mice, adding to the multi-faceted metabolic effects of GIP which need to be taken into account when developing GIPR-targeted therapies for obesity and diabetes.

Complete Genomic Assembly of Mauritian Cynomolgus Macaque Killer Ig-like Receptor and Natural Killer Group 2 Haplotypes.

Mauritian-origin cynomolgus macaques (MCMs) serve as a powerful nonhuman primate model in biomedical research due to their unique genetic homogeneity, which simplifies experimental designs. Despite their extensive use, a comprehensive understanding of crucial immune-regulating gene families, particularly killer Ig-like receptors (KIR) and NK group 2 (NKG2), has been hindered by the lack of detailed genomic reference assemblies. In this study, we employ advanced long-read sequencing techniques to completely assemble eight KIR and seven NKG2 genomic haplotypes, providing an extensive insight into the structural and allelic diversity of these immunoregulatory gene clusters. Leveraging these genomic resources, we prototype a strategy for genotyping KIR and NKG2 using short-read, whole-exome capture data, illustrating the potential for cost-effective multilocus genotyping at colony scale. These results mark a significant enhancement for biomedical research in MCMs and underscore the feasibility of broad-scale genetic investigations.

Expression of the MSPDBL2 antigen in a discrete subset of Plasmodium falciparum schizonts is regulated by GDV1 but may not be linked to sexual commitment.

The Plasmodium falciparum merozoite surface protein MSPDBL2 is a polymorphic antigen targeted by acquired immune responses, and normally expressed in only a minority of mature schizonts. The potential relationship of MSPDBL2 to sexual commitment is examined, as variable mspdbl2 transcript levels and proportions of MSPDBL2-positive mature schizonts in clinical isolates have previously correlated with levels of many sexual stage parasite gene transcripts, although not with the master regulator ap2-g. It is demonstrated that conditional overexpression of the gametocyte development protein GDV1, which promotes sexual commitment, also substantially increases the proportion of MSPDBL2-positive schizonts in culture. Conversely, truncation of the gdv1 gene is shown to prevent any expression of MSPDBL2. However, across diverse P. falciparum cultured lines, the variable proportions of MSPDBL2 positivity in schizonts do not correlate significantly with variable gametocyte conversion rates, indicating it is not involved in sexual commitment. Confirming this, examining a line with endogenous hemagglutinin-tagged AP2-G showed that the individual schizonts expressing MSPDBL2 are mostly different from those expressing AP2-G. Using a selection-linked integration system, modified P. falciparum lines were engineered to express an intact or disrupted version of MSPDBL2, showing the protein is not required for sexual commitment or early gametocyte development. Asexual parasite multiplication rates were also not affected by expression of either intact or disrupted MSPDBL2 in a majority of schizonts. Occurring alongside sexual commitment, the role of the discrete MSPDBL2-positive schizont subpopulation requires further investigation in natural infections where it is under immune selection. IMPORTANCE: Malaria parasites in the blood are remarkably variable, able to switch antigenic targets so they may survive within humans who have already developed specific immune responses. This is one of the challenges in developing vaccines against malaria. MSPDBL2 is a target of naturally acquired immunity expressed in minority proportions of schizonts, the end stages of each 2-day replication cycle in red blood cells which contain merozoites prepared to invade new red blood cells. Results show that the proportion of schizonts expressing MSPDBL2 is positively controlled by the expression of the regulatory gametocyte development protein GDV1. It was previously known that expression of GDV1 leads to increased expression of AP2-G which causes parasites to switch to sexual development, so a surprising finding here is that MSPDBL2-positive parasites are mostly distinct from those that express AP2-G. This discrete antigenic subpopulation of mostly asexual parasites is regulated alongside sexually committed parasites, potentially enabling survival under stress conditions.

Notes from Beethoven's genome.

Rapid advances over the last decade in DNA sequencing and statistical genetics enable us to investigate the genomic makeup of individuals throughout history. In a recent notable study, Begg et al.1 used Ludwig van Beethoven's hair strands for genome sequencing and explored genetic predispositions for some of his documented medical issues. Given that it was arguably Beethoven's skills as a musician and composer that made him an iconic figure in Western culture, we here extend the approach and apply it to musicality. We use this as an example to illustrate the broader challenges of individual-level genetic predictions.

Design of amyloidogenic peptide traps.

Segments of proteins with high ß-strand propensity can self-associate to form amyloid fibrils implicated in many diseases. We describe a general approach to bind such segments in ß-strand and ß-hairpin conformations using de novo designed scaffolds that contain deep peptide-binding clefts. The designs bind their cognate peptides in vitro with nanomolar affinities. The crystal structure of a designed protein-peptide complex is close to the design model, and NMR characterization reveals how the peptide-binding cleft is protected in the apo state. We use the approach to design binders to the amyloid-forming proteins transthyretin, tau, serum amyloid A1 and amyloid ß1-42 (Aß42). The Aß binders block the assembly of Aß fibrils as effectively as the most potent of the clinically tested antibodies to date and protect cells from toxic Aß42 species.

Machine learning-guided engineering of genetically encoded fluorescent calcium indicators.

Here we used machine learning to engineer genetically encoded fluorescent indicators, protein-based sensors critical for real-time monitoring of biological activity. We used machine learning to predict the outcomes of sensor mutagenesis by analyzing established libraries that link sensor sequences to functions. Using the GCaMP calcium indicator as a scaffold, we developed an ensemble of three regression models trained on experimentally derived GCaMP mutation libraries. The trained ensemble performed an in silico functional screen on 1,423 novel, uncharacterized GCaMP variants. As a result, we identified the ensemble-derived GCaMP (eGCaMP) variants, eGCaMP and eGCaMP+, which achieve both faster kinetics and larger ∆F/F0 responses upon stimulation than previously published fast variants. Furthermore, we identified a combinatorial mutation with extraordinary dynamic range, eGCaMP2+, which outperforms the tested sixth-, seventh- and eighth-generation GCaMPs. These findings demonstrate the value of machine learning as a tool to facilitate the efficient engineering of proteins for desired biophysical characteristics.

Generalized biomolecular modeling and design with RoseTTAFold All-Atom.

Deep-learning methods have revolutionized protein structure prediction and design but are presently limited to protein-only systems. We describe RoseTTAFold All-Atom (RFAA), which combines a residue-based representation of amino acids and DNA bases with an atomic representation of all other groups to model assemblies that contain proteins, nucleic acids, small molecules, metals, and covalent modifications, given their sequences and chemical structures. By fine-tuning on denoising tasks, we developed RFdiffusion All-Atom (RFdiffusionAA), which builds protein structures around small molecules. Starting from random distributions of amino acid residues surrounding target small molecules, we designed and experimentally validated, through crystallography and binding measurements, proteins that bind the cardiac disease therapeutic digoxigenin, the enzymatic cofactor heme, and the light-harvesting molecule bilin.

Computational design of soluble functional analogues of integral membrane proteins.

De novo design of complex protein folds using solely computational means remains a significant challenge. Here, we use a robust deep learning pipeline to design complex folds and soluble analogues of integral membrane proteins. Unique membrane topologies, such as those from GPCRs, are not found in the soluble proteome and we demonstrate that their structural features can be recapitulated in solution. Biophysical analyses reveal high thermal stability of the designs and experimental structures show remarkable design accuracy. The soluble analogues were functionalized with native structural motifs, standing as a proof-of-concept for bringing membrane protein functions to the soluble proteome, potentially enabling new approaches in drug discovery. In summary, we designed complex protein topologies and enriched them with functionalities from membrane proteins, with high experimental success rates, leading to a de facto expansion of the functional soluble fold space.

Rapid and automated design of two-component protein nanomaterials using ProteinMPNN.

The design of protein-protein interfaces using physics-based design methods such as Rosetta requires substantial computational resources and manual refinement by expert structural biologists. Deep learning methods promise to simplify protein-protein interface design and enable its application to a wide variety of problems by researchers from various scientific disciplines. Here, we test the ability of a deep learning method for protein sequence design, ProteinMPNN, to design two-component tetrahedral protein nanomaterials and benchmark its performance against Rosetta. ProteinMPNN had a similar success rate to Rosetta, yielding 13 new experimentally confirmed assemblies, but required orders of magnitude less computation and no manual refinement. The interfaces designed by ProteinMPNN were substantially more polar than those designed by Rosetta, which facilitated in vitro assembly of the designed nanomaterials from independently purified components. Crystal structures of several of the assemblies confirmed the accuracy of the design method at high resolution. Our results showcase the potential of deep learning-based methods to unlock the widespread application of designed protein-protein interfaces and self-assembling protein nanomaterials in biotechnology.

Blueprinting extendable nanomaterials with standardized protein blocks.

A wooden house frame consists of many different lumber pieces, but because of the regularity of these building blocks, the structure can be designed using straightforward geometrical principles. The design of multicomponent protein assemblies, in comparison, has been much more complex, largely owing to the irregular shapes of protein structures1. Here we describe extendable linear, curved and angled protein building blocks, as well as inter-block interactions, that conform to specified geometric standards; assemblies designed using these blocks inherit their extendability and regular interaction surfaces, enabling them to be expanded or contracted by varying the number of modules, and reinforced with secondary struts. Using X-ray crystallography and electron microscopy, we validate nanomaterial designs ranging from simple polygonal and circular oligomers that can be concentrically nested, up to large polyhedral nanocages and unbounded straight 'train track' assemblies with reconfigurable sizes and geometries that can be readily blueprinted. Because of the complexity of protein structures and sequence-structure relationships, it has not previously been possible to build up large protein assemblies by deliberate placement of protein backbones onto a blank three-dimensional canvas; the simplicity and geometric regularity of our design platform now enables construction of protein nanomaterials according to 'back of an envelope' architectural blueprints.

Protein Ensemble Generation Through Variational Autoencoder Latent Space Sampling.

Mapping the ensemble of protein conformations that contribute to function and can be targeted by small molecule drugs remains an outstanding challenge. Here, we explore the use of variational autoencoders for reducing the challenge of dimensionality in the protein structure ensemble generation problem. We convert high-dimensional protein structural data into a continuous, low-dimensional representation, carry out a search in this space guided by a structure quality metric, and then use RoseTTAFold guided by the sampled structural information to generate 3D structures. We use this approach to generate ensembles for the cancer relevant protein K-Ras, train the VAE on a subset of the available K-Ras crystal structures and MD simulation snapshots, and assess the extent of sampling close to crystal structures withheld from training. We find that our latent space sampling procedure rapidly generates ensembles with high structural quality and is able to sample within 1 Å of held-out crystal structures, with a consistency higher than that of MD simulation or AlphaFold2 prediction. The sampled structures sufficiently recapitulate the cryptic pockets in the held-out K-Ras structures to allow for small molecule docking.