Biological chemistry

Permanent URI for this collectionhttps://digital.lib.washington.edu/handle/1773/4896

Browse

Now showing 1 - 20 of 144

Function within Disorder: HSPB1 and HSPB5 Modulate Tau Aggregation
(2026-04-20) Cervantes, Mia; Klevit, Rachel; Nath, Abhinav
In numerous neurogenerative diseases, microtubule-associated protein tau forms fibrillar aggregates that are hallmarks of disease pathology. Small heat shock proteins (sHSPs) are ATP-independent molecular chaperones that play a critical role in maintaining protein homeostasis. Two sHSPs, HSPB1 (Hsp27) and HSPB5 (aB-crystallin), are constitutively expressed in the brain and neurons. Here, we show that HSPB1 and HSPB5 delay tau aggregation in vitro through distinct mechanisms determined by their disordered N-terminal regions (NTRs). HSPB1 inhibits tau aggregation under non-stress conditions, whereas HSPB5 requires activation by pH acidosis. Using chimeric constructs in which small NTR subregions are swapped, we identify different functional regions within the NTRs that modulate chaperone function for tau. The regionsidentified contain known sites of phosphorylation, suggesting that they are also functional control points that respond to cellular stress conditions. To further define how NTR sequence features encode function, NTR-ACD interaction network is perturbed through domain swaps, mutation, and hetero-oligomerization. Our findings support a model in which specific functional motifs within disordered NTRs of sHSPs govern activity and client preference in response to cellular stress. Finally, this thesis presents a systematic framework for organizing sHSP NTRs into five distinct Function Regions (FR-A through FR-E) across ten human sHSPs. This work also establishes experimental tools for understanding sHSP oligomerization and function.
Integrin-independent activation of Tie2 using the de novo designed protein H8mb
(2026-04-20) McCurdy, Clara; Ruohola-Baker, Hannele
The Angiopoietin–Tie2 signaling axis is a central regulator of vascular stability, yetefforts to harness it therapeutically have been limited by the poor developability and receptor promiscuity of Angiopoietin-1 (Ang1), which binds both Tie2 and α5β1 integrin. Here we report the de novo design of a compact Tie2 binding protein that overcomes these constraints. We use RFdiffusion to design a stable, high-affinity Tie2 binder that functions as a selective antagonist when monomeric, but as a potent agonist when octavalent (H8mb) that activates the Tie2/pAKT pathway without binding to α5β1 integrin. H8mb induces Tie2 activation with more rapid internalization kinetics than Ang1, suggesting that while integrin engagement is dispensable for Tie2 activation, it may modulate signaling persistence by functioning as a co-receptor for Ang1 and delaying receptor internalization. In a mouse model of Acute Respiratory Distress Syndrome (ARDS), H8mb markedly improved survival. These findings show that designed receptor binding enables dissection of co-receptor control of signaling dynamics, potentially enabling the development of more selective and effective therapeutics.
Recombinant expression of type IV pilins supports new model of pilus assembly
(2026-04-20) Applegate, Justin Nathaniel; Merz, Alexey J
Type IV pili are essential virulence factors for many pathogenic bacteria including Neisseria gonorrhoeae. Their functions include: twitching motility, biofilm formation, DNA uptake, and host cell adhesion. Several proteins at the tip of the fiber – namely PilI, PilJ, PilK, and PilC – form an initiation complex that is essential to efficient pilus formation. The exact mechanism and configuration by which these proteins assemble into the fiber has previously been unknown. We have developed a new method for recombinant expression of type IV pilin proteins. By engineering a signal peptidase I sequence into the ?-helical spine of the pilins, we are able to purify them from the periplasm without the use of detergents that could disrupt protein-protein interactions. Using this method we were able to identify an obligate heterodimer formed by PilI and PilJ. Additionally, we were able to detect PilK and the C terminal half of PilC forming a complex. These results combined with alphafold modeling and cell based assays support a model where the extreme C terminus of PilC binds to PilK as the first step of pilus assembly. This PilC-PilK dimer is then able to bind to the PilI-PilJ dimer, forming the capping complex which primes pilus extension.
Developing broad-spectrum pan-sarbecovirus countermeasures
(2026-04-20) Lee, Ji Min; Veesler, David
Two sarbecoviruses, severe acute respiratory syndrome coronavirus (SARS-CoV) in 2003 and severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) in 2019, have crossed species barriers and spilled over to humans. The latter of the two resulted in a pandemic exerting an unprecedented toll on global healthcare capacity, claiming millions of lives while simultaneously inflicting collateral damage to the world economy. Pandemic experience has deepened our understanding of sarbecoviruses and shaped our frameworks for developing therapeutics and vaccines against them. Sarbecovirus spike glycoprotein is the key molecular machinery for viral entry and fusion into the host cell, making it the principal target when developing countermeasures against sarbecoviruses. The primary goal of therapeutics and vaccines is to directly neutralize the virus and/or to confer protection by eliciting immune responses. In the following chapters of this dissertation, I detail our efforts to develop pan-sarbecovirus countermeasures with a broad spectrum. First, I will describe a de novo designed miniprotein against the SARS-CoV-2 receptor binding domain that retained its inhibitory functions across multiple SARS-CoV-2 variants. Then, I will delineate the potential of receptor tropism and zoonotic spillover of a clade 3 sarbecovirus and what countermeasures might be applicable, propelling us to prepare better for a pandemic should there be another one from sarbecoviruses. In the final sections of the dissertation, I will focus on the development and assessment of broad-spectrum pan-sarbecovirus vaccines focusing on the highly conserved prefusion-stabilized fusion machinery of sarbecoviruses. Our focus on prefusion-stabilized sarbecovirus fusion machinery vaccines provides a roadmap for the development of next-generation pan-sarbecovirus vaccines. Our work, collectively, highlights the importance of sarbecovirus countermeasures and provides a foundation for developing broadly applicable therapeutics and vaccines.
Pixel-seqV2 Enables Multi-Scale Spatial Transcriptomic Analysis of Tissue Architecture and Kidney Aging
(2026-04-20) Dong, Runze; Gu, Liangcai
Understanding how molecular programs are organized within intact tissues remainsa central challenge in biology. Conventional transcriptomics provides comprehensive molecular information but loses spatial context, whereas imaging preserves tissue architecture but lacks genome-wide molecular resolution. Spatial transcriptomics seeks to bridge this gap; however, existing platforms are constrained by trade-offs among spatial resolution, molecular sensitivity, scalability, and reproducibility. This dissertation presents the development and application of Pixel-seqV2, a scalable, sequencing-based spatial transcriptomics platform designed to enable high- density molecular capture across large tissue areas while preserving near-histological spatial fidelity. Pixel-seqV2 employs patterned flowcell–derived polony gels and stamping-based, substrate replication to achieve reproducible fabrication, tunable probe density, and continuous spatial sampling at micron-scale resolution. Systematic benchmarking demonstrates improved transcript capture efficiency, reduced lateral diffusion, and accurate recovery of RNA organization. Together with a factor-aware, Malat1-guided segmentation framework, these advances enable robust reconstruction of single-cell transcriptomes directly from sequencing-derived spatial data. Using the mouse kidney as a model system, this work illustrates how dense spatial transcriptomic measurements enable biological reasoning across multiple spatial scales. Tissue-wide analyses reconstruct nephron architecture and corticomedullary organization with near-histological clarity. At sub-glomerular resolution, Pixel-seqV2 resolves the internal organization of glomeruli and spatial relationships among interacting vascular and epithelial cell populations. Within anatomically continuous proximal tubules, localized transcriptional micro-niches characterized by Pigr- associated immune-transport programs reveal functional specialization beyond classical nephron segment boundaries. 3 Applying these approaches to aging kidneys reveals that molecular aging is not uniformly distributed across tissue but instead manifests as spatial niche remodeling. Aging is associated with coordinated loss of epithelial metabolic programs, emergence of immune-enriched cortical micro-environments, and heterogeneous decline in glomerular functional states across kidneys collected from multiple mice. These findings support a model in which tissue aging reflects reorganization of spatially localized cellular interactions rather than diffuse molecular deterioration. Together, this dissertation demonstrates that advances in spatial transcriptomic technology enable a transition from spatial measurement toward spatial reasoning. By integrating platform development, computational analysis, and biological application, Pixel-seqV2 establishes a framework for studying how complex tissues organize function, respond to stress, and remodel during aging.
Computational Design of Protein Interfaces For Therapeutic Development
(2026-04-20) Venkatesh, Preetham; Baker, David
Recent advances in deep learning have revolutionized the de novo design of protein interfaces, offering unprecedented control over structural features of binders and epitope targeting. Unlike conventional antibodies, de novo protein binders offer superior stability, ease of production, and cost-effectiveness, making them ideal for diagnostic and therapeutic applications. In Chapter 1, we address the challenge of targeting flexible bioactive helical peptides. We also introduce a novel approach to improving affinity of designs by using generative tools for dock refinement. We further demonstrate the utility of these binders as robust tools for hormone detection in biosensors and mass spectrometry. In Chapter 2, we present a generalizable platform to solubilize G-protein coupled receptors (GPCRs) by transplanting their orthosteric binding sites onto stable designed scaffolds. We demonstrate that these soluble analogs enable high-throughput screening and the discovery of novel antagonists for complex membrane targets. Together, these chapters establish design principles and computational workflows that broaden the scope of de novo binder design for drug discovery.
Transcriptomic Analysis Reveals Metabolic and Epigenomic Reprogramming During Lytic Phase of MHV-68 Infection
(2026-02-05) Jones, Chloe; Escobar, Thelma
The human gammaherpesviruses (γHV) Kaposi Sarcoma-Associated Virus (KSHV) and Epstein Barr Virus (EBV) are two of the seven known human oncogenic viruses. KSHV infects around 5% of the population and causes cancers such as Kaposi Sarcoma in immunodeficient individuals, whereas EBV infects >90% of the population and is linked to lymphomas and, more recently, multiple sclerosis. Despite their impact on human health, there are currently no drugs to treat either virus. This may in part be due to difficulties studying the lytic phase in the human γHVs; KSHV and EBV both immediately enter latency upon initial infection. A better understanding of what processes occur during the lytic phase, the part of the viral lifecycle during which the virus actively replicates, may offer new insights on potentially druggable pathways. MHV-68 is a murine γHV that immediately enters the lytic phase upon primary infection and shares ~80% sequence similarity with the human γHVs, thus making it a good model for studying the lytic phase in the human γHVs. Surprisingly, there have been no prior studies of the host transcriptome during viral infection of NIH3T3 cells, one of the most common models for the study of MHV-68. In this body of work, I analyze bulk RNA-seq data from MHV-68 infected NIH3T3 cells with the goal of identifying potential druggable pathway(s). In taking a two-pronged approach in which I assess the most differentially expressed metabolic and epigenomic genes/pathways, I present transcriptomic evidence for increased redox stress in MHV-68 infected NIH3T3 cells, as well as evidence for perturbations in chromatin biology. My analysis indicated an increase in transcripts encoding components of the mitochondrial electron transport chain, glutathione synthesis, and NADPH-producing pathways, collectively suggesting increased oxidative stress over the course of infection. Notable changes in chromatin biology include upregulation of the expression of DNA demethylases and the DNA damage response chaperone DAXX. Additionally, canonical histone H2A was upregulated, while histone variant H2AX was downregulated late in lytic infection. Ultimately, I found that inhibition of the DNA demethylating enzymes Ten-eleven translocase 1 and 2 (Tet1 and Tet2) diminishes MHV-68 replication in this system. The mechanism behind how Tet1/2 inhibition reduces viral titer remains to be explored.
Engineering novel nanomaterials through de novo design of hydrophobic scaffold proteins
(2026-02-05) Zambrano, Dane; King, Neil
Computationally designed protein nanoparticles leverage de novo designed protein subunits to build large oligomers which can be used for multivalent antigen display and cargo loading in the delivery of vaccines. Advancements in computational methods and design of membrane proteins increases the breadth of materials which can be used as subunit building blocks in these large oligomeric complexes. In this dissertation I describe engineering a protein scaffold with a hydrophobic pore and the rational design of a novel two-component nanoparticle utilizing a de novo designed transmembrane protein. Using machine learning and AI-guided approaches with two-component RPX Docking, AlphaFold 2 and RosettaFold Diffusion, I structurally characterized a novel two component transmembrane nanoparticle and to 4.16 Å by Cryo-EM. This technology marks the first generation of novel hydrophobic nanoparticles that is a step towards generating in vitro hybrid lipid-protein nanomaterials for the display of unique membrane proteins and lipid-conjugated moieties previously inaccessible through current designed nanoparticles.
Characterizing the Structural and Physiological Effects of IMPDH2 Mutations Associated with Neurodevelopmental Disorders
(2025-10-02) O'Neill, Audrey; Kollman, Justin M
Inosine-5'-monophosphate dehydrogenase (IMPDH) catalyzes the first committed step of de novo guanine nucleotide biosynthesis, converting IMP to XMP. To control this important metabolic branch point between adenine and guanine nucleotide synthesis, IMPDH is highly regulated, including through assembly into filaments. There are two isoforms of IMPDH in humans, but IMPDH2 is specifically essential for development and is upregulated during proliferation. Mutations in IMPDH2 have been identified in patients with neurodevelopmental disorders exhibiting a range of neurological symptoms, including dystonia. Here, we show with in vitro enzyme assays, negative stain electron microscopy, and high-resolution structures determined by cryo-EM, how each mutation affects the structure, activity, and allosteric regulation of IMPDH2 filaments and octamers. We also develop Xenopus tropicalis as a model to study the effects of one variant, the in-frame deletion of serine 160, on metabolism, neuromuscular development, and IMPDH filament formation in a vertebrate system. This work establishes a model for studying the mechanisms of disease that arise from IMPDH2 dysregulation.
Sequence-Specific DNA-binding Proteins: Protein Design, Structure Prediction, and Binding Prediction
(2025-10-02) McHugh, Lilian; DiMaio, Frank; Baker, David
Sequence-specific DNA-binding proteins (DBPs) perform critical roles in biology and biotechnology, and have seen decades of effort to engineer, predict, and understand their functions. In this work, I present methods to design novel DBPs, predict the structures of protein-DNA complexes, and predict the binding specificities of structurally diverse DBPs. Made with custom computational methods, we screened over 100,000 designed DBPs and identified 44 that bound their intended targets with high affinity. Several of the designed DBPs are highly specific for their targets, as demonstrated by all-by-all cross-reactivity studies, mutation-scanning competition assays, and protein-binding microarrays. The designed DBPs bind consistently with their design models, as determined via interface ablation studies and crystallographic structure determination. For structure prediction, I developed and tested RosettaFold-NA (RFNA), the first end-to-end trained machine learning model that predicts the structures of any combination of protein and nucleic acids. RFNA accurately predicts about 30% of protein-nucleic acid complexes without sequence homology to the training set. For binding prediction, I curated a dataset of over 3000 DBPs with semi-manually-assigned DNA-binding domains and hundreds of thousands of corresponding experimentally-verified DNA target sequences. I fine-tuned RFNA and RoseTTAFold-allatom both on prediction of a binary binding / non-binding classification task and on prediction of distilled protein-DNA complex structures. In a retrospective analysis of design results, the resulting fine-tuned model is able to enrich for functional designed DBPs. Using a simulated annealing inference approach, the fine-tuned model can also predict DNA-binding profiles for validation set transcription factors with reasonable accuracy and efficiency.
Computational and Rational Stabilization of Toll-Like Receptors for the Development of Novel Tools
(2025-08-01) Criswell, Cameron Ann; King, Neil
Toll-like receptors (TLRs) are membrane-bound pattern recognition receptors essential for innate immune sensing, but their large, glycosylated extracellular domains and intrinsic instability make them notoriously difficult to express and purify recombinantly. These challenges have historically limited structural and functional studies, as well as the development of therapeutic reagents targeting TLRs. This dissertation presents a computational design framework for stabilizing and functionally interrogating TLRs, enabling the development of synthetic immunomodulatory tools. Using AI-guided design with ProteinMPNN, AlphaFold2 and RosettaFold Diffusion in addition to physics and rational approaches, we generated stabilized and expressible variants of TLR2 and TLR5, facilitating downstream applications including de novo minibinder generation and antibody development. Overall, this work highlights a generalizable approach to stabilizing immune receptors and advancing rational immunotherapy design.
Mitochondrial dynamics and purine metabolism in zebrafish cone photoreceptors
(2025-08-01) Rutter, Kaitlyn Michelle; Brockerhoff, Susan E
Photoreceptors, the light-detecting cells in the retina, are vital for vision. Despite being post-mitotic and having high energy demands, photoreceptors survive throughout a human’s lifespan. However, many different proteins, including inosine monophosphate dehydrogenase 1 (IMPDH1), can cause vision loss when mutated. In this thesis, I describe a strategy used by zebrafish photoreceptors to remain healthy throughout their lifespan. I also present new hypotheses to explain why cone photoreceptors degenerate when faced with a genetically-induced metabolic insult. Although photoreceptors are mostly glycolytic, they have abundant mitochondria that help with energy demand. We found that cone photoreceptor mitochondria are dynamic depending on the time of day and energy demands. Further, we found that at night, cone mitochondria extrude material out of cones. We expanded our studies to evaluate how larval zebrafish cone mitochondria handle stress. In stressed and unstressed conditions, cone mitochondria migrate away from the ellipsoid region. These unhealthy mitochondria are ejected from the cell, and MÃ¼ller glia take them up for turnover. We also tested a metabolic stressor by expressing mutant forms of Impdh1a in cone photoreceptors. In humans, these mutations leadto photoreceptor degeneration. Some Impdh1a mutations are hyperactive in vitro so we designed the first Impdh1a mutant animal models to analyze enzyme activity in vivo. We found no evidence for Impdh1a hyperactivity in zebrafish cone photoreceptors and instead found other metabolic phenotypes that could contribute to disease. We also found that Impdh1a mutants form large mislocalized filaments that may disrupt important protein-protein interactions. These studies are important for understanding cone health in the context of daily energy demand changes, stress, and disease.
Development of Neural Networks for Biomolecular Structure Prediction with Applications to Protein Design
(2025-08-01) Krishna, Rohith; Baker, David
A grand challenge in biology is to create computational models of the interactions betweenabitrary biomolecular structures. In this dissertation, I describe the development of neural network models for predicting the structure of biomolecular complexes including proteins, nucleic acids, and small molecules. First, we developed a general neural network architecture for the prediction of biomolecular complexes in the Protein Data Bank (PDB). We then demonstrated the ability of this model to predict the structure of new complexes with high accuracy. Subsequently, we applied this model of native biomolecular complexes to the design of de novo small molecule binding proteins and enzymes. Finally, we developed a framework for development of future neural networks trained on the PDB and apply it to train several structure prediction models. To our knowledge, this dissertation represents the first efforts to develop general-purpose neural network models for biomolecular structure prediction and design.
The Regulation of Mfn1-Mediated Mitochondrial Outer Membrane Fusion
(2025-05-12) Martinez Bocanegra, Jennifer; Hoppins, Suzanne
Mitochondria are often depicted in textbooks as static organelles whose principal purpose is to power cells; however, they are constantly moving and morphing shape in response to cellular requirements which go beyond the energetic needs of the cell. The dynamic properties exhibited by mitochondria include transport and positioning via interactions with cytoskeleton, and fission and fusion of the mitochondrial membranes, all of which are critical for maintaining mitochondrial function. The implications of disrupted mitochondrial transport and division have been extensively described in comparison to disrupted mitochondrial fusion. However, both mitochondrial fission and fusion are required to maintain a balanced and healthy network of mitochondria able to effectively respond to cellular demands. Mitochondrial fission and fusion are mediated by large GTPases from the Dynamin Superfamily of Proteins. Mitochondrial outer membrane fusion is mediated in mammals by the paralogs Mfn1 and Mfn2, collectively known as the mitofusins. Mutations in MFN2 are the cause for the most common form of Charcot-Marie-Tooth Type 2A Disease (CMT2A), though the mechanistic relationship between dysfunction of mitochondrial outer membrane fusion and the presentation of symptoms in patients with CMT2A has not been previously investigated. The following work describes the use of two biochemically characterized Mfn1 variants, Mfn1 S329P and Mfn1 S228A, to further investigate how mitofusin function is regulated and how regulation of Mfn1 contributes to cellular function. First, I present the use of the nematode Caenorhabditis elegans as a powerful new model to understand the contribution of Mfn1 S329P to neuronal dysfunction. This variant, along with its homologous disease-associated variant Mfn2 S350P, are fusion incompetent and exhibit phenotypes when expressed in cells that include redistribution of mitochondria to the perinuclear region. Mfn1 S329P was expressed in C. elegans and neuronal function was assessed by measuring neuron-dependent behaviors. Transgenic worms expressing Mfn1 S329P exhibited uncoordinated locomotion and decreased progeny that was likely due to an egg-laying defect. To better understand how mitofusin dysfunction contributes to disease, I performed genetic screens for suppressors of the behavioral phenotypes. Given that a candidate-based approach did not yield insight into the cellular functions contributing to the behavioral defects, I performed an unbiased genetic screen to identify candidates. I isolated agns-1 (um0017) in C. elegans and describe the formation of Mfn1 S329P associated mitochondrial clusters in neurons, which is facilitated by agns-1. Finally, I propose leveraging C. elegans as a tractable model for investigating how CMT2A-associated mitofusin variants contribute to neuronal dysfunction in real-time. In the following chapter I investigate the functional impacts of phosphorylation at Mfn1 S228 with respect to cellular health. I find that blocking phosphorylation of Mfn1 S228 renders cells slightly more susceptible to cell death induced by the ATP synthase inhibitor Oligomycin. Taken together, I demonstrate that dysregulation of Mfn1 contributes to cellular dysfunction and provide a novel model that bridges the gap between our understanding of mitofusin function in vitro.
A Biophysical and Structural Assessment of the Spillover Potential of Bat-Borne Sarbecoviruses
(2025-01-23) Zepeda, Samantha Kathleen; Veesler, David J.
The propensity of coronaviruses to spillover from animal reservoirs into human populations has proven to have a startling impact on society in recent years. During the first two outbreaks, caused by severe acute respiratory syndrome coronavirus (SARS-CoV) in 2002-2003 and Middle East respiratory syndrome coronavirus (MERS-CoV) in 2012-present, death tolls hovered around 800 each (World Health Organization, 2004, 2019). During the recent emergence of SARS-CoV-2 in 2019, previous death tolls were shattered as SARS-CoV-2 tore across the globe; the death toll to date stands in the millions. These viruses all began in their native reservoirs, bats, some spread to intermediate hosts palm civets (SARS-CoV-1) and dromedary camels (MERS-CoV), but all started in bats and eventually made their way into humans. They all like all other coronaviruses utilize their spike glycoproteins to mediate viral attachment and fusion to host cells. During each of these spillover events, we were unprepared, we lacked vaccines or therapeutics, and we knew very little about what existed in animal reservoirs. Today, the situation is very much changed, vaccines and therapeutics have been developed against SARS-CoV-2, but there is still an urgent need to understand what these viruses are doing in their hosts, to identify the barriers to spillover, and characterize and stockpile vaccines and therapeutics for future spillover preparedness.In the following dissertation, I detail the deep ancestral origins of ACE2 receptor usage in sarbecoviruses and describe the evolvability and plasticity of ACE2 binding in sarbecoviruses. I also highlight sarbecoviruses currently unable to bind and utilize human ACE2 for cellular entry but with the capacity to gain human ACE2 binding and cellular entry within 1-2 amino acid mutations. I then describe a Cambodian bat sarbecovirus, closely related to SARS-CoV-2, with the capacity to bind human ACE2 but unable to facilitate cellular entry and explore the role of spike conformational dynamics on the spillover potential of this virus. I further describe a bat-borne sarbecovirus that is much more distally related to SARS-CoV-2, PRD-0038, originating from African and explore its receptor usage, antigenicity, and immunogenicity; demonstrating that vaccination with this spike glycoprotein results in broader vaccine mismatch protection, and motivating the inclusion of clade 3 antigens in next-generation vaccines for enhanced resilience to viral evolution. I detail the discovery of ACE2 utilization in HKU5 and describe molecular basis for multiple independent evolution of ACE2 across coronaviruses. I also detail our effort towards pandemic response and zoonotic spillover preparedness; describing the discovery and characterization of several broadly neutralizing antibodies, the elicitation of broad neutralizing antibodies with multivalent RBD-nanoparticle vaccines and describe the impact of mutation of SARS-CoV-2 Omicron variants on receptor usage, fusogenicity, and immune/therapeutic evasion. Finally, I detail the development of a ligand fitting software EMERALD, and its usage in fitting linoleic acid in spike glycoproteins and other protein datasets. Collectively, my work informs on barriers to spillover, the molecular mechanisms of coronavirus receptor usage, the evolution of sarbecoviruses in their natural reservoirs, and on the development of the next generation of vaccines and therapeutics.
Targeting arbitrary regions of intrinsically disordered proteins
(2025-01-23) Wu, Kejia; Baker, David
A general, robust approach to design proteins that bind tightly and specifically to intrinsically disordered regions (IDRs) of proteins and flexible peptides with minimal cost would have wide applications in biological research, therapeutics, and diagnosis. However, the lack of defined structures and the high variability in sequence and conformational preferences has complicated such efforts. Herein, we have built layers of work to solve this problem with two main computational methods, i.e., bottom-up and top-down. As for the bottom-up approach, we built a landscape to first develop components of geometric matching and amino acid sidechain-oriented pocket recognition for regular peptides; we then generalized it to the broad non-regular landscapes combined with deep learning tools under a rule of optimal binding geometric constraints, achieving specific recognition toward arbitrary unstructured protein sequence space. As for the top-down approach, we leveraged the power of deep learning, trained neural networks to predict and co-fold a disordered target and a designed binding protein to it all together. Using these computational methods, we have designed binders to more than 50 broadly diverse unstructured targets, including highly polar targets. Experimental testing of dozens to hundreds of designs per target yielded binders with affinities better than 100 nM in most cases, and in the pM range straight out of the computer in five cases. Co-crystal structures of designed binder-target complexes as well as NMR structures with isotope labeled peptide targets were closely consistent with the design models. All-by-all in vitro binding crosstalk experiments for representative designs binding diverse targets show they are highly specific for the intended targets, with no crosstalk even for the closely related peptides. Designs were shown functional in a number of downstream assays indicating the therapeutic, diagnosis, intracellular tracking potentials. These methods were applied in the biologically relevant cancer target RAS to distinguish the four distinct isoforms in cells to the degree antibodies have never achieved. Our approach thus could provide a general solution to the intrinsically disordered protein and peptide recognition problem, while paving a road to site-specific recognition of post-translational modifications (PTMs) and enzymatic functional designs as well.
Genetically Encoded Optical Biosensors for Drugs and Peptides Comprising De Novo Engineered Synthetic Mini-binders
(2025-01-23) Pan, Yuxin; Gu, Liangcai
Methods for de novo engineering of protein binders into chemically induced dimerization (CID) systems offer new opportunities for developing genetically encoded sensors for drugs and metabolites that lack suitable natural binders. However, integrating these binders into genetically encoded fluorescent sensors (GEFSs) remains largely unexplored. Here, we present a pipeline that efficiently selects synthetic CIDs and integrates them with a fluorescent domain to create single-protein GEFSs suitable for solution-based assays and mammalian cell applications.As proof of concept, we used this pipeline to generate synthetic CID pairs from monobody and nanobody scaffolds and incorporated them into circularly permuted green fluorescent protein-based GEFSs. To optimize sensor performance, we created a library of 361 unique linker variants and conducted lysate-based screening to identify the most effective configurations. The resulting sensors exhibited a ΔF/F₀ exceeding 100%. Furthermore, these sensors retained functionality in HEK293T cells when localized to the plasma membrane, cytoplasm, or ER lumen. We also adapted this approach to additional optical domains, such as circularly permuted HaloTag , demonstrating its broad applicability. Our results establish a foundation for expanding synthetic GEFSs, significantly enhancing the small molecule and peptide GEFS toolkit and reducing reliance on natural proteins and derivatives.
Computational design of serine hydrolases
(2025-01-23) Lauko, Anna; Baker, David
Nature’s enzymes are exceptionally powerful catalysts, exerting dramatic rate accelerations and exquisite control over a remarkable variety of chemical transformations. Since their initial discovery and characterization, the ability to generate artificial enzymes for chemical reactions involved in industrial processes, chemical synthesis, and therapeutic applications has been of considerable interest. Despite decades of effort, artificial enzymes continue to display lower catalytic activities than their native counterparts, even for well-understood model reactions. Here, we present a novel and general approach to computational enzyme design utilizing recent advances in tailored protein scaffold generation and active site conformational ensemble prediction. As a proof of concept, we have applied this method to the design of esterases that utilize the serine hydrolase enzymatic mechanism. Despite a deep understanding of the mechanism amassed through decades of study, previous attempts to design esterases acting through this mechanism have failed. To our knowledge, the designs made using our approach represent the first examples of accurately designed, de novo serine hydrolases spanning folds not found in natural hydrolases and exhibiting catalytic efficiencies on par with hydrolases in nature that act on similar substrates. We believe our approach will not only enable the design of industrially relevant serine hydrolases but also be broadly applicable to accelerating a wider array of chemical reactions, including ones that do not occur in nature.
De novo design of protein binders to bioactive peptides: from hormones to snake toxins
(2024-10-16) Vazquez Torres, Susana; Baker, David D.
Recent advancements in deep learning methods for protein design and structure prediction have demonstrated the potential of de novo proteins to replace antibodies as affinity reagents, offering cost-effective production, increased stability, and robust platforms for therapeutics and diagnostics. In this work, we set out to design de novo protein binders for two classes of bioactive peptides: hormones which adopt α-helical structures, and lethal snake toxins from the three-finger toxin (3FTX) family. Using deep learning-based methods, we generated high-affinity binders for both targets. These binders enable hormone detection via mass spectrometry and the creation of bioluminescence-based biosensors. Additionally, our designed proteins effectively neutralize 3FTXs in vitro and protect mice from lethal neurotoxin challenges. While further development is necessary to translate these designs into practical diagnostics and therapeutics, this work demonstrates their potential as robust reagents for hormone detection and as safer, widely accessible antivenoms.
Molecular Basis of Plasticity of Porcine Delta Coronavirus (PDCoV) & Development of Countermeasures Against Infection
(2024-10-16) Rexhepaj, Megi; Kollman, Justin M
Porcine deltacoronavirus (PDCoV) is an enteric pathogen that infects a broad range of mammal and avian species. Viral entry is achieved through the transmembrane spike (S) glycoprotein, specifically the receptor binding domain (RBD), binding to host receptor aminopeptidase N (APN). The S glycoprotein plays a key role in modulating host and tissue tropism, zoonotic transmission, and pathogenesis. The mechanism by which the S glycoprotein binds to a broad range of host receptors is not understood. Given that the S glycoprotein is the main target of antibodies and that neutralizing antibody titers are a correlate of protection against coronaviruses, the development of vaccines and therapeutics focuses intensively on this glycoprotein target. This research describes the molecular basis of binding of the S glycoprotein to host receptors and identifies the neutralizing epitopes on the S glycoproteins providing the first line of protection for a possible future PDCoV epidemic.

Browse

Recent Submissions