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.

A role for vessel-associated extracellular matrix proteins in multiple sclerosis pathology.

Multiple sclerosis (MS) is unsurpassed for its clinical and pathological hetherogeneity, but the biological determinants of this variability are unknown. HLA-DRB1*15, the main genetic risk factor for MS, influences the severity and distribution of MS pathology. This study set out to unravel the molecular determinants of the heterogeneity of MS pathology in relation to HLA-DRB1*15 status. Shotgun proteomics from a discovery cohort of MS spinal cord samples segregated by HLA-DRB*15 status revealed overexpression of the extracellular matrix (ECM) proteins, biglycan, decorin, and prolargin in HLA-DRB*15-positive cases, adding to established literature on a role of ECM proteins in MS pathology that has heretofore lacked systematic pathological validation. These findings informed a neuropathological characterisation of these proteins in a large autopsy cohort of 41 MS cases (18 HLA-DRB1*15-positive and 23 HLA-DRB1*15-negative), and seven non-neurological controls on motor cortical, cervical and lumbar spinal cord tissue. Biglycan and decorin demonstrate a striking perivascular expression pattern in controls that is reduced in MS (-36.5%, p = 0.036 and - 24.7%, p = 0.039; respectively) in lesional and non-lesional areas. A concomitant increase in diffuse parenchymal accumulation of biglycan and decorin is seen in MS (p = 0.015 and p = 0.001, respectively), particularly in HLA-DRB1*15-positive cases (p = 0.007 and p = 0.046, respectively). Prolargin shows a faint parenchymal pattern in controls that is markedly increased in MS cases where a perivascular deposition pattern is observed (motor cortex +97.5%, p = 0.001; cervical cord +49.1%, p = 0.016). Our findings point to ECM proteins and the vascular interface playing a central role in MS pathology within and outside the plaque area. As ECM proteins are known potent pro-inflammatory molecules, their parenchymal accumulation may contribute to disease severity. This study brings to light novel factors that may contribute to the heterogeneity of the topographical variation of MS pathology.

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.

Protein Design Using Structure-Prediction Networks: AlphaFold and RoseTTAFold as Protein Structure Foundation Models.

Designing proteins with tailored structures and functions is a long-standing goal in bioengineering. Recently, deep learning advances have enabled protein structure prediction at near-experimental accuracy, which has catalyzed progress in protein design as well. We review recent studies that use structure-prediction neural networks to design proteins, via approaches such as activation maximization, inpainting, or denoising diffusion. These methods have led to major improvements over previous methods in wet-lab success rates for designing protein binders, metalloproteins, enzymes, and oligomeric assemblies. These results show that structure-prediction models are a powerful foundation for developing protein-design tools and suggest that continued improvement of their accuracy and generality will be key to unlocking the full potential of protein design.

De novo design of high-affinity binders of bioactive helical peptides.

Many peptide hormones form an α-helix on binding their receptors1-4, and sensitive methods for their detection could contribute to better clinical management of disease5. De novo protein design can now generate binders with high affinity and specificity to structured proteins6,7. However, the design of interactions between proteins and short peptides with helical propensity is an unmet challenge. Here we describe parametric generation and deep learning-based methods for designing proteins to address this challenge. We show that by extending RFdiffusion8 to enable binder design to flexible targets, and to refining input structure models by successive noising and denoising (partial diffusion), picomolar-affinity binders can be generated to helical peptide targets by either refining designs generated with other methods, or completely de novo starting from random noise distributions without any subsequent experimental optimization. The RFdiffusion designs enable the enrichment and subsequent detection of parathyroid hormone and glucagon by mass spectrometry, and the construction of bioluminescence-based protein biosensors. The ability to design binders to conformationally variable targets, and to optimize by partial diffusion both natural and designed proteins, should be broadly useful.

HEATR5B associates with dynein-dynactin and promotes motility of AP1-bound endosomal membranes.

The microtubule motor dynein mediates polarised trafficking of a wide variety of organelles, vesicles and macromolecules. These functions are dependent on the dynactin complex, which helps recruit cargoes to dynein's tail and activates motor movement. How the dynein-dynactin complex orchestrates trafficking of diverse cargoes is unclear. Here, we identify HEATR5B, an interactor of the adaptor protein-1 (AP1) clathrin adaptor complex, as a novel player in dynein-dynactin function. HEATR5B was recovered in a biochemical screen for proteins whose association with the dynein tail is augmented by dynactin. We show that HEATR5B binds directly to the dynein tail and dynactin and stimulates motility of AP1-associated endosomal membranes in human cells. We also demonstrate that the Drosophila HEATR5B homologue is an essential gene that selectively promotes dynein-based transport of AP1-bound membranes to the Golgi apparatus. As HEATR5B lacks the coiled-coil architecture typical of dynein adaptors, our data point to a non-canonical process orchestrating motor function on a specific cargo. We additionally show that HEATR5B promotes association of AP1 with endosomal membranes independently of dynein. Thus, HEATR5B co-ordinates multiple events in AP1-based trafficking.

Macromolecular condensation buffers intracellular water potential.

Optimum protein function and biochemical activity critically depends on water availability because solvent thermodynamics drive protein folding and macromolecular interactions1. Reciprocally, macromolecules restrict the movement of 'structured' water molecules within their hydration layers, reducing the available 'free' bulk solvent and therefore the total thermodynamic potential energy of water, or water potential. Here, within concentrated macromolecular solutions such as the cytosol, we found that modest changes in temperature greatly affect the water potential, and are counteracted by opposing changes in osmotic strength. This duality of temperature and osmotic strength enables simple manipulations of solvent thermodynamics to prevent cell death after extreme cold or heat shock. Physiologically, cells must sustain their activity against fluctuating temperature, pressure and osmotic strength, which impact water availability within seconds. Yet, established mechanisms of water homeostasis act over much slower timescales2,3; we therefore postulated the existence of a rapid compensatory response. We find that this function is performed by water potential-driven changes in macromolecular assembly, particularly biomolecular condensation of intrinsically disordered proteins. The formation and dissolution of biomolecular condensates liberates and captures free water, respectively, quickly counteracting thermal or osmotic perturbations of water potential, which is consequently robustly buffered in the cytoplasm. Our results indicate that biomolecular condensation constitutes an intrinsic biophysical feedback response that rapidly compensates for intracellular osmotic and thermal fluctuations. We suggest that preserving water availability within the concentrated cytosol is an overlooked evolutionary driver of protein (dis)order and function.

Zero-shot mutation effect prediction on protein stability and function using RoseTTAFold.

Predicting the effects of mutations on protein function and stability is an outstanding challenge. Here, we assess the performance of a variant of RoseTTAFold jointly trained for sequence and structure recovery, RFjoint , for mutation effect prediction. Without any further training, we achieve comparable accuracy in predicting mutation effects for a diverse set of protein families using RFjoint to both another zero-shot model (MSA Transformer) and a model that requires specific training on a particular protein family for mutation effect prediction (DeepSequence). Thus, although the architecture of RFjoint was developed to address the protein design problem of scaffolding functional motifs, RFjoint acquired an understanding of the mutational landscapes of proteins during model training that is equivalent to that of recently developed large protein language models. The ability to simultaneously reason over protein structure and sequence could enable even more precise mutation effect predictions following supervised training on the task. These results suggest that RFjoint has a quite broad understanding of protein sequence-structure landscapes, and can be viewed as a joint model for protein sequence and structure which could be broadly useful for protein modeling.

Synthetic Par polarity induces cytoskeleton asymmetry in unpolarized mammalian cells.

Polarized cells rely on a polarized cytoskeleton to function. Yet, how cortical polarity cues induce cytoskeleton polarization remains elusive. Here, we capitalized on recently established designed 2D protein arrays to ectopically engineer cortical polarity of virtually any protein of interest during mitosis in various cell types. This enables direct manipulation of polarity signaling and the identification of the cortical cues sufficient for cytoskeleton polarization. Using this assay, we dissected the logic of the Par complex pathway, a key regulator of cytoskeleton polarity during asymmetric cell division. We show that cortical clustering of any Par complex subunit is sufficient to trigger complex assembly and that the primary kinetic barrier to complex assembly is the relief of Par6 autoinhibition. Further, we found that inducing cortical Par complex polarity induces two hallmarks of asymmetric cell division in unpolarized mammalian cells: spindle orientation, occurring via Par3, and central spindle asymmetry, depending on aPKC activity.

De novo design of protein structure and function with RFdiffusion.

There has been considerable recent progress in designing new proteins using deep-learning methods1-9. Despite this progress, a general deep-learning framework for protein design that enables solution of a wide range of design challenges, including de novo binder design and design of higher-order symmetric architectures, has yet to be described. Diffusion models10,11 have had considerable success in image and language generative modelling but limited success when applied to protein modelling, probably due to the complexity of protein backbone geometry and sequence-structure relationships. Here we show that by fine-tuning the RoseTTAFold structure prediction network on protein structure denoising tasks, we obtain a generative model of protein backbones that achieves outstanding performance on unconditional and topology-constrained protein monomer design, protein binder design, symmetric oligomer design, enzyme active site scaffolding and symmetric motif scaffolding for therapeutic and metal-binding protein design. We demonstrate the power and generality of the method, called RoseTTAFold diffusion (RFdiffusion), by experimentally characterizing the structures and functions of hundreds of designed symmetric assemblies, metal-binding proteins and protein binders. The accuracy of RFdiffusion is confirmed by the cryogenic electron microscopy structure of a designed binder in complex with influenza haemagglutinin that is nearly identical to the design model. In a manner analogous to networks that produce images from user-specified inputs, RFdiffusion enables the design of diverse functional proteins from simple molecular specifications.

ANXA11 biomolecular condensates facilitate protein-lipid phase coupling on lysosomal membranes.

Phase transitions of cellular proteins and lipids play a key role in governing the organisation and coordination of intracellular biology. The frequent juxtaposition of proteinaceous biomolecular condensates to cellular membranes raises the intriguing prospect that phase transitions in proteins and lipids could be co-regulated. Here we investigate this possibility in the ribonucleoprotein (RNP) granule-ANXA11-lysosome ensemble, where ANXA11 tethers RNP granule condensates to lysosomal membranes to enable their co-trafficking. We show that changes to the protein phase state within this system, driven by the low complexity ANXA11 N-terminus, induce a coupled phase state change in the lipids of the underlying membrane. We identify the ANXA11 interacting proteins ALG2 and CALC as potent regulators of ANXA11-based phase coupling and demonstrate their influence on the nanomechanical properties of the ANXA11-lysosome ensemble and its capacity to engage RNP granules. The phenomenon of protein-lipid phase coupling we observe within this system offers an important template to understand the numerous other examples across the cell whereby biomolecular condensates closely juxtapose cell membranes.

Scaffolding protein functional sites using deep learning.

The binding and catalytic functions of proteins are generally mediated by a small number of functional residues held in place by the overall protein structure. Here, we describe deep learning approaches for scaffolding such functional sites without needing to prespecify the fold or secondary structure of the scaffold. The first approach, "constrained hallucination," optimizes sequences such that their predicted structures contain the desired functional site. The second approach, "inpainting," starts from the functional site and fills in additional sequence and structure to create a viable protein scaffold in a single forward pass through a specifically trained RoseTTAFold network. We use these two methods to design candidate immunogens, receptor traps, metalloproteins, enzymes, and protein-binding proteins and validate the designs using a combination of in silico and experimental tests.

Compensatory ion transport buffers daily protein rhythms to regulate osmotic balance and cellular physiology.

Between 6-20% of the cellular proteome is under circadian control and tunes mammalian cell function with daily environmental cycles. For cell viability, and to maintain volume within narrow limits, the daily variation in osmotic potential exerted by changes in the soluble proteome must be counterbalanced. The mechanisms and consequences of this osmotic compensation have not been investigated before. In cultured cells and in tissue we find that compensation involves electroneutral active transport of Na+, K+, and Cl- through differential activity of SLC12A family cotransporters. In cardiomyocytes ex vivo and in vivo, compensatory ion fluxes confer daily variation in electrical activity. Perturbation of soluble protein abundance has commensurate effects on ion composition and cellular function across the circadian cycle. Thus, circadian regulation of the proteome impacts ion homeostasis with substantial consequences for the physiology of electrically active cells such as cardiomyocytes.

High-efficacy subcellular micropatterning of proteins using fibrinogen anchors.

Protein micropatterning allows proteins to be precisely deposited onto a substrate of choice and is now routinely used in cell biology and in vitro reconstitution. However, drawbacks of current technology are that micropatterning efficiency can be variable between proteins and that proteins may lose activity on the micropatterns. Here, we describe a general method to enable micropatterning of virtually any protein at high specificity and homogeneity while maintaining its activity. Our method is based on an anchor that micropatterns well, fibrinogen, which we functionalized to bind to common purification tags. This enhances micropatterning on various substrates, facilitates multiplexed micropatterning, and dramatically improves the on-pattern activity of fragile proteins like molecular motors. Furthermore, it enhances the micropatterning of hard-to-micropattern cells. Last, this method enables subcellular micropatterning, whereby complex micropatterns simultaneously control cell shape and the distribution of transmembrane receptors within that cell. Altogether, these results open new avenues for cell biology.

Design of biologically active binary protein 2D materials.

Ordered two-dimensional arrays such as S-layers1,2 and designed analogues3-5 have intrigued bioengineers6,7, but with the exception of a single lattice formed with flexible linkers8, they are constituted from just one protein component. Materials composed of two components have considerable potential advantages for modulating assembly dynamics and incorporating more complex functionality9-12. Here we describe a computational method to generate co-assembling binary layers by designing rigid interfaces between pairs of dihedral protein building blocks, and use it to design a p6m lattice. The designed array components are soluble at millimolar concentrations, but when combined at nanomolar concentrations, they rapidly assemble into nearly crystalline micrometre-scale arrays nearly identical to the computational design model in vitro and in cells without the need for a two-dimensional support. Because the material is designed from the ground up, the components can be readily functionalized and their symmetry reconfigured, enabling formation of ligand arrays with distinguishable surfaces, which we demonstrate can drive extensive receptor clustering, downstream protein recruitment and signalling. Using atomic force microscopy on supported bilayers and quantitative microscopy on living cells, we show that arrays assembled on membranes have component stoichiometry and structure similar to arrays formed in vitro, and that our material can therefore impose order onto fundamentally disordered substrates such as cell membranes. In contrast to previously characterized cell surface receptor binding assemblies such as antibodies and nanocages, which are rapidly endocytosed, we find that large arrays assembled at the cell surface suppress endocytosis in a tunable manner, with potential therapeutic relevance for extending receptor engagement and immune evasion. Our work provides a foundation for a synthetic cell biology in which multi-protein macroscale materials are designed to modulate cell responses and reshape synthetic and living systems.

Quantitative Chemical Delivery of Quantum Dots into the Cytosol of Cells.

The ability to image single molecules in living cells has been impaired by the absence of bright, photostable fluorophores. Quantum dots (QDs) offer an attractive solution to this problem due to their exceptional photostability and brightness. Here, we describe in detail a protocol to chemically deliver functionalized QDs into the cytosol of living cells, based on cell-penetrating poly(disulfide)s (CPDs). This protocol is highly efficient and delivers hundreds of QDs per cell after incubation of cells with functionalized QDs at nanomolar concentrations. We also detail a pipeline for automated detection and tracking of diffusive QDs in living cells, which may provide a useful means to study the biophysical properties of the cytosol and their dynamics. Last, we describe a protocol for conjugating streptavidin fusion proteins to QDs, in order to permit the codelivery of QDs with functional proteins of interest into cells. The protocol has been successfully applied to a broad range of different cell types, thus offering a flexible and generalizable means to image single molecules in living cells.