High-throughput characterization of protein-protein interactions by reprogramming yeast mating

RESP2: An uncertainty aware multi-target multi-property optimization AI pipeline for antibody discovery

Discovery of therapeutic antibodies against infectious disease pathogens presents distinct challenges. Ideal candidates must possess not only the properties required for any therapeutic antibody (e.g. specificity, low immunogenicity) but also high affinity to many mutants of the target antigen. Here we present RESP2, an enhanced version of our RESP pipeline, designed for the discovery of antibodies against diverse antigens with simultaneously optimized developability properties. RESP2 provides a suite of methods to estimate the uncertainty of predictions including a new model combining neural network and Gaussian process with great flexibility to model protein engineering data, which accelerates in silico directed evolution to identify tight binders even those not present in the original screening library. An interpretable model is then exploited to assess antibody humanness to minimize immunogenicity risk of the selected candidates. To demonstrate the power of this pipeline, we use the receptor binding domain (RBD) of the COVID-19 spike protein as a case study, and discover a highly human antibody with broad (mid to high-affinity) binding to at least 8 different variants of the RBD. These results illustrate the advantages of this pipeline for antibody discovery against a challenging target. The code needed to reproduce the experiments in this paper is available at https://github.com/Wang-lab-UCSD/RESP2.

Mammalian cell display with automated oligo design and library assembly allows for rapid residue level conformational epitope mapping

Precise epitope determination of therapeutic antibodies is of great value as it allows for further comprehension of mechanism of action, therapeutic responsiveness prediction, avoidance of unwanted cross reactivity, and vaccine design. The golden standard for discontinuous epitope determination is the laborious X-ray crystallography method. Here, we present a combinatorial method for rapid mapping of discontinuous epitopes by mammalian antigen display, eliminating the need for protein expression and purification. The method is facilitated by automated workflows and tailored software for antigen analysis and oligonucleotide design. These oligos are used in automated mutagenesis to generate an antigen receptor library displayed on mammalian cells for direct binding analysis by flow cytometry. Through automated analysis of 33930 primers an optimized single condition cloning reaction was defined allowing for mutation of all surface-exposed residues of the receptor binding domain of SARS-CoV-2. All variants were functionally expressed, and two reference binders validated the method. Furthermore, epitopes of three novel therapeutic antibodies were successfully determined followed by evaluation of binding also towards SARS-CoV-2 Omicron BA.2. We find the method to be highly relevant for rapid construction of antigen libraries and determination of antibody epitopes, especially for the development of therapeutic interventions against novel pathogens.

A library-on-library screen reveals the breadth expansion landscape of a broadly neutralizing betacoronavirus antibody

Broadly neutralizing antibodies (bnAbs) typically evolve cross-reactivity breadth through acquiring somatic hypermutations. While evolution of breadth requires improvement of binding to multiple antigenic variants, most experimental evolution platforms select against only one antigenic variant at a time. In this study, a yeast display library-on-library approach was applied to delineate the affinity maturation of a betacoronavirus bnAb, S2P6, against 27 spike stem helix peptides in a single experiment. Our results revealed that the binding affinity landscape of S2P6 varies among different stem helix peptides. However, somatic hypermutations that confer general improvement in binding affinity across different stem helix peptides could also be identified. We further showed that a key somatic hypermutation for breadth expansion involves long-range interaction. Overall, our work not only provides a proof-of-concept for using a library-on-library approach to analyze the evolution of antibody breadth, but also has important implications for the development of broadly protective vaccines.

A Computational Pipeline for Accurate Prioritization of Protein-Protein Binding Candidates in High-Throughput Protein Libraries

Identifying the interactome for a protein of interest is challenging due to the large number of possible binders. High-throughput experimental approaches narrow down possible binding partners but often include false positives. Furthermore, they provide no information about what the binding region is (e.g., the binding epitope). We introduce a novel computational pipeline based on an AlphaFold2 (AF) Competitive Binding Assay (AF-CBA) to identify proteins that bind a target of interest from a pull-down experiment and the binding epitope. Our focus is on proteins that bind the Extraterminal (ET) domain of Bromo and Extraterminal domain (BET) proteins, but we also introduce nine additional systems to show transferability to other peptide-protein systems. We describe a series of limitations to the methodology based on intrinsic deficiencies of AF and AF-CBA to help users identify scenarios where the approach will be most useful. Given the method's speed and accuracy, we anticipate its broad applicability to identify binding epitope regions among potential partners, setting the stage for experimental verification.

Systematic genetic characterization of the human PKR kinase domain highlights its functional malleability to escape a viral substrate mimic

Evolutionary arms races can arise at the contact surfaces between host and viral proteins, producing dynamic spaces in which genetic variants are continually pursued. However, the sampling of genetic variation must be balanced with the need to maintain protein function. A striking case is given by protein kinase R (PKR), a member of the mammalian innate immune system. PKR detects viral replication within the host cell and halts protein synthesis to prevent viral replication by phosphorylating eIF2α, a component of the translation initiation machinery. PKR is targeted by many viral antagonists, including poxvirus pseudosubstrate antagonists that inhibit PKR by interacting with the same binding surface as eIF2α. Remarkably, PKR has several rapidly evolving residues at this interface, suggesting it is engaging in an evolutionary arms race, despite the surface's critical role in phosphorylating eIF2α. To systematically explore the evolutionary opportunities available at this dynamic interface, we generated and characterized a library of 426 SNP-accessible nonsynonymous variants of human PKR for their ability to escape inhibition by the model pseudosubstrate inhibitor K3 from vaccinia virus. We identified key sites in the PKR kinase domain that harbor K3-resistant variants, as well as critical sites where variation leads to loss of function. We find K3-resistant variants are readily available throughout the interface and are enriched at sites under positive selection. Moreover, variants beneficial against K3 were also beneficial against an enhanced variant of K3, indicating resilience to viral adaptation. Overall, we find that the eIF2α-binding surface of PKR is highly malleable, potentiating its evolutionary ability to combat viral inhibition.

Yeast-based screening platforms to understand and improve human health

Detailed molecular understanding of the human organism is essential to develop effective therapies. Saccharomyces cerevisiae has been used extensively for acquiring insights into important aspects of human health, such as studying genetics and cell-cell communication, elucidating protein-protein interaction (PPI) networks, and investigating human G protein-coupled receptor (hGPCR) signaling. We highlight recent advances and opportunities of yeast-based technologies for cost-efficient chemical library screening on hGPCRs, accelerated deciphering of PPI networks with mating-based screening and selection, and accurate cell-cell communication with human immune cells. Overall, yeast-based technologies constitute an important platform to support basic understanding and innovative applications towards improving human health.

Recent advances in generative biology for biotherapeutic discovery

Generative biology combines artificial intelligence (AI), advanced life sciences technologies, and automation to revolutionize the process of designing novel biomolecules with prescribed properties, giving drug discoverers the ability to escape the limitations of biology during the design of next-generation protein therapeutics. Significant hurdles remain, namely: (i) the inherently complex nature of drug discovery, (ii) the bewildering number of promising computational and experimental techniques that have emerged in the past several years, and (iii) the limited availability of relevant protein sequence-function data for drug-like molecules. There is a need to focus on computational methods that will be most practically effective for protein drug discovery and on building experimental platforms to generate the data most appropriate for these methods. Here, we discuss recent advances in computational and experimental life sciences that are most crucial for impacting the pace and success of protein drug discovery.

Building Representation Learning Models for Antibody Comprehension

Antibodies are versatile proteins with both the capacity to bind a broad range of targets and a proven track record as some of the most successful therapeutics. However, the development of novel antibody therapeutics is a lengthy and costly process. It is challenging to predict the functional and biophysical properties of antibodies from their amino acid sequence alone, requiring numerous experiments for full characterization. Machine learning, specifically deep representation learning, has emerged as a family of methods that can complement wet lab approaches and accelerate the overall discovery and engineering process. Here, we review advances in antibody sequence representation learning, and how this has improved antibody structure prediction and facilitated antibody optimization. We discuss challenges in the development and implementation of such models, such as the lack of publicly available, well-curated antibody function data and highlight opportunities for improvement. These and future advances in machine learning for antibody sequences have the potential to increase the success rate in developing new therapeutics, resulting in broader access to transformative medicines and improved patient outcomes.

Sifting Through the Noise: A Computational Pipeline for Accurate Prioritization of Protein-Protein Binding Candidates in High-Throughput Protein Libraries

Identifying the interactome for a protein of interest is challenging due to the large number of possible binders. High-throughput experimental approaches narrow down possible binding partners, but often include false positives. Furthermore, they provide no information about what the binding region is (e.g. the binding epitope). We introduce a novel computational pipeline based on an AlphaFold2 (AF) Competition Assay (AF-CBA) to identify proteins that bind a target of interest from a pull-down experiment, along with the binding epitope. Our focus is on proteins that bind the Extraterminal (ET) domain of Bromo and Extraterminal domain (BET) proteins, but we also introduce nine additional systems to show transferability to other peptide-protein systems. We describe a series of limitations to the methodology based on intrinsic deficiencies to AF and AF-CBA, to help users identify scenarios where the approach will be most useful. Given the speed and accuracy of the methodology, we expect it to be generally applicable to facilitate target selection for experimental verification starting from high-throughput protein libraries.

Soil microbiome engineering for sustainability in a changing environment

Recent advances in microbial ecology and synthetic biology have the potential to mitigate damage caused by anthropogenic activities that are deleteriously impacting Earth's soil ecosystems. Here, we discuss challenges and opportunities for harnessing natural and synthetic soil microbial communities, focusing on plant growth promotion under different scenarios. We explore current needs for microbial solutions in soil ecosystems, how these solutions are being developed and applied, and the potential for new biotechnology breakthroughs to tailor and target microbial products for specific applications. We highlight several scientific and technological advances in soil microbiome engineering, including characterization of microbes that impact soil ecosystems, directing how microbes assemble to interact in soil environments, and the developing suite of gene-engineering approaches. This Review underscores the need for an interdisciplinary approach to understand the composition, dynamics and deployment of beneficial soil microbiomes to drive efforts to mitigate or reverse environmental damage by restoring and protecting healthy soil ecosystems.

Therapies for Tau-associated neurodegenerative disorders: targeting molecules, synapses, and cells

Advances in experimental and computational technologies continue to grow rapidly to provide novel avenues for the treatment of neurodegenerative disorders. Despite this, there remain only a handful of drugs that have shown success in late-stage clinical trials for Tau-associated neurodegenerative disorders. The most commonly prescribed treatments are symptomatic treatments such as cholinesterase inhibitors and N-methyl-D-aspartate receptor blockers that were approved for use in Alzheimer's disease. As diagnostic screening can detect disorders at earlier time points, the field needs pre-symptomatic treatments that can prevent, or significantly delay the progression of these disorders (Koychev et al., 2019). These approaches may be different from late-stage treatments that may help to ameliorate symptoms and slow progression once symptoms have become more advanced should early diagnostic screening fail. This mini-review will highlight five key avenues of academic and industrial research for identifying therapeutic strategies to treat Tau-associated neurodegenerative disorders. These avenues include investigating (1) the broad class of chemicals termed "small molecules"; (2) adaptive immunity through both passive and active antibody treatments; (3) innate immunity with an emphasis on microglial modulation; (4) synaptic compartments with the view that Tau-associated neurodegenerative disorders are synaptopathies. Although this mini-review will focus on Alzheimer's disease due to its prevalence, it will also argue the need to target other tauopathies, as through understanding Alzheimer's disease as a Tau-associated neurodegenerative disorder, we may be able to generalize treatment options. For this reason, added detail linking back specifically to Tau protein as a direct therapeutic target will be added to each topic.

Massively parallel protein-protein interaction measurement by sequencing (MP3-seq) enables rapid screening of protein heterodimers

Protein-protein interactions (PPIs) regulate many cellular processes, and engineered PPIs have cell and gene therapy applications. Here we introduce massively parallel protein-protein interaction measurement by sequencing (MP3-seq), an easy-to-use and highly scalable yeast-two-hybrid approach for measuring PPIs. In MP3-seq, DNA barcodes are associated with specific protein pairs, and barcode enrichment can be read by sequencing to provide a direct measure of interaction strength. We show that MP3-seq is highly quantitative and scales to over 100,000 interactions. We apply MP3-seq to characterize interactions between families of rationally designed heterodimers and to investigate elements conferring specificity to coiled-coil interactions. Finally, we predict coiled heterodimer structures using AlphaFold-Multimer (AF-M) and train linear models on physics simulation energy terms to predict MP3-seq values. We find that AF-M and AF-M complex prediction-based models could be valuable for pre-screening interactions, but that measuring interactions experimentally remains necessary to rank their strengths quantitatively.

A multiplexed bacterial two-hybrid for rapid characterization of protein-protein interactions and iterative protein design

Protein-protein interactions (PPIs) are crucial for biological functions and have applications ranging from drug design to synthetic cell circuits. Coiled-coils have been used as a model to study the sequence determinants of specificity. However, building well-behaved sets of orthogonal pairs of coiled-coils remains challenging due to inaccurate predictions of orthogonality and difficulties in testing at scale. To address this, we develop the next-generation bacterial two-hybrid (NGB2H) method, which allows for the rapid exploration of interactions of programmed protein libraries in a quantitative and scalable way using next-generation sequencing readout. We design, build, and test large sets of orthogonal synthetic coiled-coils, assayed over 8,000 PPIs, and used the dataset to train a more accurate coiled-coil scoring algorithm (iCipa). After characterizing nearly 18,000 new PPIs, we identify to the best of our knowledge the largest set of orthogonal coiled-coils to date, with fifteen on-target interactions. Our approach provides a powerful tool for the design of orthogonal PPIs.

Deploying synthetic coevolution and machine learning to engineer protein-protein interactions

Fine-tuning of protein-protein interactions occurs naturally through coevolution, but this process is difficult to recapitulate in the laboratory. We describe a platform for synthetic protein-protein coevolution that can isolate matched pairs of interacting muteins from complex libraries. This large dataset of coevolved complexes drove a systems-level analysis of molecular recognition between Z domain-affibody pairs spanning a wide range of structures, affinities, cross-reactivities, and orthogonalities, and captured a broad spectrum of coevolutionary networks. Furthermore, we harnessed pretrained protein language models to expand, in silico, the amino acid diversity of our coevolution screen, predicting remodeled interfaces beyond the reach of the experimental library. The integration of these approaches provides a means of simulating protein coevolution and generating protein complexes with diverse molecular recognition properties for biotechnology and synthetic biology.

Machine learning optimization of candidate antibody yields highly diverse sub-nanomolar affinity antibody libraries

Therapeutic antibodies are an important and rapidly growing drug modality. However, the design and discovery of early-stage antibody therapeutics remain a time and cost-intensive endeavor. Here we present an end-to-end Bayesian, language model-based method for designing large and diverse libraries of high-affinity single-chain variable fragments (scFvs) that are then empirically measured. In a head-to-head comparison with a directed evolution approach, we show that the best scFv generated from our method represents a 28.7-fold improvement in binding over the best scFv from the directed evolution. Additionally, 99% of designed scFvs in our most successful library are improvements over the initial candidate scFv. By comparing a library's predicted success to actual measurements, we demonstrate our method's ability to explore tradeoffs between library success and diversity. Results of our work highlight the significant impact machine learning models can have on scFv development. We expect our method to be broadly applicable and provide value to other protein engineering tasks.

The next wave of interactomics: Mapping the SLiM-based interactions of the intrinsically disordered proteome

Short linear motifs (SLiMs) are a unique and ubiquitous class of protein interaction modules that perform key regulatory functions and drive dynamic complex formation. For decades, interactions mediated by SLiMs have accumulated through detailed low-throughput experiments. Recent methodological advances have opened this previously underexplored area of the human interactome to high-throughput protein-protein interaction discovery. In this article, we discuss that SLiM-based interactions represent a significant blind spot in the current interactomics data, introduce the key methods that are illuminating the elusive SLiM-mediated interactome of the human cell on a large scale, and discuss the implications for the field.

Titrating Avidity of Yeast-Displayed Proteins Using a Transcriptional Regulator

Yeast surface display is a valuable tool for protein engineering and directed evolution; however, significant variability in the copy number (i.e., avidity) of displayed variants on the yeast cell wall complicates screening and selection campaigns. Here, we report an engineered titratable display platform that modulates the avidity of Aga2-fusion proteins on the yeast cell wall dependent on the concentration of the anhydrotetracycline (aTc) inducer. Our design is based on a genomic Aga1 gene copy and an episomal Aga2-fusion construct both under the control of an aTc-dependent transcriptional regulator that enables stoichiometric and titratable expression, secretion, and display of Aga2-fusion proteins. We demonstrate tunable display levels over 2-3 orders of magnitude for various model proteins, including glucose oxidase enzyme variants, mechanostable dockerin-binding domains, and anti-PDL1 affibody domains. By regulating the copy number of displayed proteins, we demonstrate the effects of titratable avidity levels on several specific phenotypic activities, including enzyme activity and cell adhesion to surfaces under shear flow. Finally, we show that titrating down the display level allows yeast-based binding affinity measurements to be performed in a regime that avoids ligand depletion effects while maintaining small sample volumes, avoiding a well-known artifact in yeast-based binding assays. The ability to titrate the multivalency of proteins on the yeast cell wall through simple inducer control will benefit protein engineering and directed evolution methodology relying on yeast display for broad classes of therapeutic and diagnostic proteins of interest.

A focus on yeast mating: From pheromone signaling to cell-cell fusion

Cells live in a chemical environment and are able to orient towards chemical cues. Unicellular haploid fungal cells communicate by secreting pheromones to reproduce sexually. In the yeast models Saccharomyces cerevisiae and Schizosaccharomyces pombe, pheromonal communication activates similar pathways composed of cognate G-protein-coupled receptors and downstream small GTPase Cdc42 and MAP kinase cascades. Local pheromone release and sensing, at a mobile surface polarity patch, underlie spatial gradient interpretation to form pairs between two cells of distinct mating types. Concentration of secretion at the point of cell-cell contact then leads to local cell wall digestion for cell fusion, forming a diploid zygote that prevents further fusion attempts. A number of asymmetries between mating types may promote efficiency of the system. In this review, we present our current knowledge of pheromone signaling in the two model yeasts, with an emphasis on how cells decode the pheromone signal spatially and ultimately fuse together. Though overall pathway architectures are similar in the two species, their large evolutionary distance allows to explore how conceptually similar solutions to a general biological problem can arise from divergent molecular components.

Bioinformatic Analysis of Yeast Two-Hybrid Next-Generation Interaction Screen Data

Yeast two-hybrid next-generation interaction screening (Y2H-NGIS) uses the output of next-generation sequencing to mine for novel protein-protein interactions. Here, we outline the analytics underlying Y2H-NGIS datasets. Different systems, libraries, and experimental designs comprise Y2H-NGIS methodologies. We summarize the analysis in several layers that comprise the characterization of baits and preys, quantification, and identification of true interactions for subsequent secondary validation. We present two software designed for this purpose, NGPINT and Y2H-SCORES, which are used as front-end and back-end tools in the analysis. Y2H-SCORES software can be used and adapted to analyze different datasets not only from Y2H-NGIS but from other techniques ruled by similar biological principles.

A dataset comprised of binding interactions for 104,972 antibodies against a SARS-CoV-2 peptide

The dataset presented here contains quantitative binding scores of scFv-format antibodies against a SARS-CoV-2 target peptide collected via an AlphaSeq assay that can be used in the development and benchmarking of machine learning models. Starting from three seed sequences identified from a phage display campaign using a human naïve library, four sets of 29,900 antibodies were designed in silico by creating all k = 1 mutations and random k = 2 and k = 3 mutations throughout the complementary-determining regions (CDRs). Of the 119,600 designs, 104,972 were successfully built in to the AlphaSeq library and target binding was subsequently measured with 71,384 designs resulting in a predicted affinity value for at least one of the triplicate measurements. Data include antibodies with predicted affinity measurements ranging from 37 pM to 22 mM. To our knowledge, this dataset is the largest, publicly available dataset that contains antibody sequences, antigen sequence and quantitative measurements of binding scores and provides an opportunity to serve as a benchmark to evaluate antibody-specific representation models for machine learning.

Measurement(s)	Antibody Binding
Technology Type(s)	AlphaSeq
Factor Type(s)	Antibody sequence
Sample Characteristic - Organism	Homo sapiens

helixCAM: A platform for programmable cellular assembly in bacteria and human cells

Interactions between cells are indispensable for signaling and creating structure. The ability to direct precise cell-cell interactions would be powerful for engineering tissues, understanding signaling pathways, and directing immune cell targeting. In humans, intercellular interactions are mediated by cell adhesion molecules (CAMs). However, endogenous CAMs are natively expressed by many cells and tend to have cross-reactivity, making them unsuitable for programming specific interactions. Here, we showcase "helixCAM," a platform for engineering synthetic CAMs by presenting coiled-coil peptides on the cell surface. helixCAMs were able to create specific cell-cell interactions and direct patterned aggregate formation in bacteria and human cells. Based on coiled-coil interaction principles, we built a set of rationally designed helixCAM libraries, which led to the discovery of additional high-performance helixCAM pairs. We applied this helixCAM toolkit for various multicellular engineering applications, such as spherical layering, adherent cell targeting, and surface patterning.

A Modified Yeast Two-Hybrid Platform Enables Dynamic Control of Expression Intensities to Unmask Properties of Protein-Protein Interactions

The yeast two-hybrid (Y2H) assay is widely used for protein-protein interaction characterization due to its simplicity and accessibility. However, it may mask changes in affinity caused by mutations or ligand activation due to signal saturation. To overcome this drawback, we modified the Y2H system to have tunable protein expression by introducing a fluorescent reporter and a pair of synthetic inducible transcription factors to regulate the expression of interacting components. We found that the application of inducers allowed us to adjust the concentrations of interacting proteins to avoid saturation and observe interactions otherwise masked in the canonical Y2H assay, such as the abscisic acid-mediated increase in affinity of monomeric abscisic acid receptors to the coreceptor. When applied in future studies, our modified system may provide a more accurate characterization of protein-protein interactions.

Massively Multiplexed Affinity Characterization of Therapeutic Antibodies Against SARS-CoV-2 Variants

Antibody therapies represent a valuable tool to reduce COVID-19 deaths and hospitalizations. Multiple antibody candidates have been granted emergency use authorization by the FDA and many more are in clinical trials. Most antibody therapies for COVID-19 are engineered to bind to the receptor-binding domain (RBD) of the SARS-CoV-2 Spike protein and disrupt its interaction with ACE2. Notably, several SARS-CoV-2 strains have accrued mutations throughout the RBD that improve ACE2 binding affinity, enhance viral transmission, and escape some existing antibody therapies. Here, we measure the binding affinity of 33 therapeutic antibodies against a large panel of SARS-CoV-2 variants and related strains of clinical significance using AlphaSeq, a high-throughput yeast mating based assay to determine epitopic residues, determine which mutations result in loss of binding, and predict how future RBD variants may impact antibody efficacy. We find that some antibodies are more susceptible to mutations in the RBD than others, for example Imdevimab and Sotrovimab display no loss in binding to B.1.1.7, B.1.351 and P.1 variants whereas we observe a 10-fold loss in binding of Etesevimab and Regdanvimab to these variants.

Rapid screening of TCR-pMHC interactions by the YAMTAD system

Personalized immunotherapy, such as cancer vaccine and TCR-T methods, demands rapid screening of TCR-pMHC interactions. While several screening approaches have been developed, their throughput is limited. Here, the Yeast Agglutination Mediated TCR antigen Discovery system (YAMTAD) was designed and demonstrated to allow fast and unbiased library-on-library screening of TCR-pMHC interactions. Our proof-of-principle study achieved high sensitivity and specificity in identifying antigens for a given TCR and identifying TCRs recognizing a given pMHC for modest library sizes. Finally, the enrichment of high-affinity TCR-pMHC interactions by YAMTAD in library-on-library screening was demonstrated. Given the high throughput (10⁶–10⁸ × 10⁶–10⁸ in theory) and simplicity (identifying TCR-pMHC interactions without purification of TCR and pMHC) of YAMTAD, this study provides a rapid but effective platform for TCR-pMHC interaction screening, with valuable applications in future personalized immunotherapy.

Antigen identification and high-throughput interaction mapping by reprogramming viral entry

Deciphering immune recognition is critical for understanding a broad range of diseases and for the development of effective vaccines and immunotherapies. Efforts to do so are limited by a lack of technologies capable of simultaneously capturing the complexity of adaptive immunoreceptor repertoires and the landscape of potential antigens. To address this, we present receptor-antigen pairing by targeted retroviruses, which combines viral pseudotyping and molecular engineering approaches to enable one-pot library-on-library interaction screens by displaying antigens on the surface of lentiviruses and encoding their identity in the viral genome. Antigen-specific viral infection of cell lines expressing human T or B cell receptors allows readout of both antigen and receptor identities via single-cell sequencing. The resulting system is modular, scalable and compatible with any cell type. These techniques provide a suite of tools for targeted viral entry, molecular engineering and interaction screens with broad potential applications.

Titration of apparent in-cellula affinities of protein-protein interactions

A genetic assay permits simultaneous quantification of two interacting proteins and their bound fraction at the single-cell level using flow cytometry. Apparent in-cellula affinities of protein-protein interactions can be extracted from the acquired data through a titration-like analysis. The applicability of this approach is demonstrated on a diverse set of interactions with proteins from different families and organisms and with in-vitro dissociation constants ranging from picomolar to micromolar.

Directed evolution of orthogonal RNA-RBP pairs through library-vs-library in vitro selection

RNA-binding proteins (RBPs) and their RNA ligands play many critical roles in gene regulation and RNA processing in cells. They are also useful for various applications in cell biology and synthetic biology. However, re-engineering novel and orthogonal RNA-RBP pairs from natural components remains challenging while such synthetic RNA-RBP pairs could significantly expand the RNA-RBP toolbox for various applications. Here, we report a novel library-vs-library in vitro selection strategy based on Phage Display coupled with Systematic Evolution of Ligands by EXponential enrichment (PD-SELEX). Starting with pools of 1.1 × 1012 unique RNA sequences and 4.0 × 108 unique phage-displayed L7Ae-scaffold (LS) proteins, we selected RNA-RBP complexes through a two-step affinity purification process. After six rounds of library-vs-library selection, the selected RNAs and LS proteins were analyzed by next-generation sequencing (NGS). Further deconvolution of the enriched RNA and LS protein sequences revealed two synthetic and orthogonal RNA-RBP pairs that exhibit picomolar affinity and >4000-fold selectivity.

A genome-scale yeast library with inducible expression of individual genes

The ability to switch a gene from off to on and monitor dynamic changes provides a powerful approach for probing gene function and elucidating causal regulatory relationships. Here, we developed and characterized YETI (Yeast Estradiol strains with Titratable Induction), a collection in which > 5,600 yeast genes are engineered for transcriptional inducibility with single-gene precision at their native loci and without plasmids. Each strain contains SGA screening markers and a unique barcode, enabling high-throughput genetics. We characterized YETI using growth phenotyping and BAR-seq screens, and we used a YETI allele to identify the regulon of Rof1, showing that it acts to repress transcription. We observed that strains with inducible essential genes that have low native expression can often grow without inducer. Analysis of data from eukaryotic and prokaryotic systems shows that native expression is a variable that can bias promoter-perturbing screens, including CRISPRi. We engineered a second expression system, Z3 EB42, that gives lower expression than Z3 EV, a feature enabling conditional activation and repression of lowly expressed essential genes that grow without inducer in the YETI library.

Applications of Machine and Deep Learning in Adaptive Immunity

Adaptive immunity is mediated by lymphocyte B and T cells, which respectively express a vast and diverse repertoire of B cell and T cell receptors and, in conjunction with peptide antigen presentation through major histocompatibility complexes (MHCs), can recognize and respond to pathogens and diseased cells. In recent years, advances in deep sequencing have led to a massive increase in the amount of adaptive immune receptor repertoire data; additionally, proteomics techniques have led to a wealth of data on peptide-MHC presentation. These large-scale data sets are now making it possible to train machine and deep learning models, which can be used to identify complex and high-dimensional patterns in immune repertoires. This article introduces adaptive immune repertoires and machine and deep learning related to biological sequence data and then summarizes the many applications in this field, which span from predicting the immunological status of a host to the antigen specificity of individual receptors and the engineering of immunotherapeutics.

Next-generation yeast-two-hybrid analysis with Y2H-SCORES identifies novel interactors of the MLA immune receptor

Protein-protein interaction networks are one of the most effective representations of cellular behavior. In order to build these models, high-throughput techniques are required. Next-generation interaction screening (NGIS) protocols that combine yeast two-hybrid (Y2H) with deep sequencing are promising approaches to generate interactome networks in any organism. However, challenges remain to mining reliable information from these screens and thus, limit its broader implementation. Here, we present a computational framework, designated Y2H-SCORES, for analyzing high-throughput Y2H screens. Y2H-SCORES considers key aspects of NGIS experimental design and important characteristics of the resulting data that distinguish it from RNA-seq expression datasets. Three quantitative ranking scores were implemented to identify interacting partners, comprising: 1) significant enrichment under selection for positive interactions, 2) degree of interaction specificity among multi-bait comparisons, and 3) selection of in-frame interactors. Using simulation and an empirical dataset, we provide a quantitative assessment to predict interacting partners under a wide range of experimental scenarios, facilitating independent confirmation by one-to-one bait-prey tests. Simulation of Y2H-NGIS enabled us to identify conditions that maximize detection of true interactors, which can be achieved with protocols such as prey library normalization, maintenance of larger culture volumes and replication of experimental treatments. Y2H-SCORES can be implemented in different yeast-based interaction screenings, with an equivalent or superior performance than existing methods. Proof-of-concept was demonstrated by discovery and validation of novel interactions between the barley nucleotide-binding leucine-rich repeat (NLR) immune receptor MLA6, and fourteen proteins, including those that function in signaling, transcriptional regulation, and intracellular trafficking.

Animal Immunization, in Vitro Display Technologies, and Machine Learning for Antibody Discovery

For years, a discussion has persevered on the benefits and drawbacks of antibody discovery using animal immunization versus in vitro selection from non-animal-derived recombinant repertoires using display technologies. While it has been argued that using recombinant display libraries can reduce animal consumption, we hold that the number of animals used in immunization campaigns is dwarfed by the number sacrificed during preclinical studies. Thus, improving quality control of antibodies before entering in vivo studies will have a larger impact on animal consumption. Both animal immunization and recombinant repertoires present unique advantages for discovering antibodies that are fit for purpose. Furthermore, we anticipate that machine learning will play a significant role within discovery workflows, refining current antibody discovery practices.

Quantitative Yeast-Yeast Two Hybrid for the Discovery and Binding Affinity Estimation of Protein-Protein Interactions

Quantifying the binding affinity of protein-protein interactions is important for elucidating connections within biochemical signaling pathways, as well as characterization of binding proteins isolated from combinatorial libraries. We describe a quantitative yeast-yeast two-hybrid (qYY2H) system that not only enables the discovery of specific protein-protein interactions but also efficient, quantitative estimation of their binding affinities (K_D). In qYY2H, the bait and prey proteins are expressed as yeast cell surface fusions using yeast surface display. We developed a semiempirical framework for estimating the K_D of monovalent bait-prey interactions, using measurements of bait-prey yeast-yeast binding, which is mediated by multivalent interactions between yeast-displayed bait and prey. Using qYY2H, we identified interaction partners of SMAD3 and the tandem WW domains of YAP from a cDNA library and characterized their binding affinities. Finally, we showed that qYY2H could also quantitatively evaluate binding interactions mediated by post-translational modifications on the bait protein.

Aquarium: open-source laboratory software for design, execution and data management

Automation has been shown to improve the replicability and scalability of biomedical and bioindustrial research. Although the work performed in many labs is repetitive and can be standardized, few academic labs can afford the time and money required to automate their workflows with robotics. We propose that human-in-the-loop automation can fill this critical gap. To this end, we present Aquarium, an open-source, web-based software application that integrates experimental design, inventory management, protocol execution and data capture. We provide a high-level view of how researchers can install Aquarium and use it in their own labs. We discuss the impacts of the Aquarium on working practices, use in biofoundries and opportunities it affords for collaboration and education in life science laboratory research and manufacture.

Enzymatic Analysis of Yeast Cell Wall-Resident GAPDH and Its Secretion

In yeast, many proteins are found in both the cytoplasmic and extracellular compartments, and consequently it can be difficult to distinguish nonconventional secretion from cellular leakage. Therefore, we monitored the extracellular glyceraldehyde-3-phosphate dehydrogenase (GAPDH) activity of intact cells as a specific marker for nonconventional secretion. Extracellular GAPDH activity was proportional to the number of cells assayed, increased with incubation time, and was dependent on added substrates. Preincubation of intact cells with 100 μM dithiothreitol increased the reaction rate, consistent with increased access of the enzyme after reduction of cell wall disulfide cross-links. Such treatment did not increase cell permeability to propidium iodide, in contrast to effects of higher concentrations of reducing agents. An amine-specific membrane-impermeant biotinylation reagent specifically inactivated extracellular GAPDH. The enzyme was secreted again after a 30- to 60-min lag following the inactivation, and there was no concomitant increase in propidium iodide staining. There were about 4 × 10⁴ active GAPDH molecules per cell at steady state, and secretion studies showed replenishment to that level 1 h after inactivation. These results establish conditions for specific quantitative assays of cell wall proteins in the absence of cytoplasmic leakage and for subsequent quantification of secretion rates in intact cells.IMPORTANCE Eukaryotic cells secrete many proteins, including many proteins that do not follow the classical secretion pathway. Among these, the glycolytic enzyme glyceraldehyde-3-phosphate dehydrogenase (GAPDH) is unexpectedly found in the walls of yeasts and other fungi and in extracellular space in mammalian cell cultures. It is difficult to quantify extracellular GAPDH, because leakage of just a little of the very large amount of cytoplasmic enzyme can invalidate the determinations. We used enzymatic assays of intact cells while also maintaining membrane integrity. The results lead to estimates of the amount of extracellular enzyme and its rate of secretion to the wall in intact cells. Therefore, enzyme assays under controlled conditions can be used to investigate nonconventional secretion more generally.

Probing Biomolecular Interactions by a Pattern-Forming Peptide-Conjugate Sensor

As a key mechanism underpinning many biological processes, protein self-organization has been extensively studied. However, the potential to apply the distinctive, nonlinear biochemical properties of such self-organizing systems to biotechnological problems such as the facile detection and characterization of biomolecular interactions has not yet been explored. Here, we describe an in vitro assay in a 96-well plate format that harnesses the emergent behavior of the Escherichia coli Min system to provide a readout of biomolecular interactions. Crucial for the development of our approach is a minimal MinE-derived peptide that stimulates MinD ATPase activity only when dimerized. We found that this behavior could be induced via any pair of foreign, mutually binding molecular entities fused to the minimal MinE peptide. The resulting MinD ATPase activity and the spatiotemporal nature of the produced protein patterns quantitatively correlate with the affinity of the fused binding partners, thereby enabling a highly sensitive assay for biomolecular interactions. Our assay thus provides a unique means of quantitatively visualizing biomolecular interactions and may prove useful for the assessment of domain interactions within protein libraries and for the facile investigation of potential inhibitors of protein–protein interactions.

In vivo mRNA display enables large-scale proteomics by next generation sequencing

Large-scale proteomic methods are essential for the functional characterization of proteins in their native cellular context. However, proteomics has lagged far behind genomic approaches in scalability, standardization, and cost. Here, we introduce in vivo mRNA display, a technology that converts a variety of proteomics applications into a DNA sequencing problem. In vivo-expressed proteins are coupled with their encoding messenger RNAs (mRNAs) via a high-affinity stem-loop RNA binding domain interaction, enabling high-throughput identification of proteins with high sensitivity and specificity by next generation DNA sequencing. We have generated a high-coverage in vivo mRNA display library of the Saccharomyces cerevisiae proteome and demonstrated its potential for characterizing subcellular localization and interactions of proteins expressed in their native cellular context. In vivo mRNA display libraries promise to circumvent the limitations of mass spectrometry-based proteomics and leverage the exponentially improving cost and throughput of DNA sequencing to systematically characterize native functional proteomes.

Automated design of thousands of nonrepetitive parts for engineering stable genetic systems

Engineered genetic systems are prone to failure when their genetic parts contain repetitive sequences. Designing many nonrepetitive genetic parts with desired functionalities remains a difficult challenge with high computational complexity. To overcome this challenge, we developed the Nonrepetitive Parts Calculator to rapidly generate thousands of highly nonrepetitive genetic parts from specified design constraints, including promoters, ribosome-binding sites and terminators. As a demonstration, we designed and experimentally characterized 4,350 nonrepetitive bacterial promoters with transcription rates that varied across a 820,000-fold range, and 1,722 highly nonrepetitive yeast promoters with transcription rates that varied across a 25,000-fold range. We applied machine learning to explain how specific interactions controlled the promoters' transcription rates. We also show that using nonrepetitive genetic parts substantially reduces homologous recombination, resulting in greater genetic stability.

A Quantitative Tri-fluorescent Yeast Two-hybrid System: From Flow Cytometry to In cellula Affinities

We present a technological advancement for the estimation of the affinities of Protein-Protein Interactions (PPIs) in living cells. A novel set of vectors is introduced that enables a quantitative yeast two-hybrid system based on fluorescent fusion proteins. The vectors allow simultaneous quantification of the reaction partners (Bait and Prey) and the reporter at the single-cell level by flow cytometry. We validate the applicability of this system on a small but diverse set of PPIs (eleven protein families from six organisms) with different affinities; the dissociation constants range from 117 pm to 17 μm After only two hours of reaction, expression of the reporter can be detected even for the weakest PPI. Through a simple gating analysis, it is possible to select only cells with identical expression levels of the reaction partners. As a result of this standardization of expression levels, the mean reporter levels directly reflect the affinities of the studied PPIs. With a set of PPIs with known affinities, it is straightforward to construct an affinity ladder that permits rapid classification of PPIs with thus far unknown affinities. Conventional software can be used for this analysis. To permit automated analysis, we provide a graphical user interface for the Python-based FlowCytometryTools package.

A Phage-Assisted Continuous Selection Approach for Deep Mutational Scanning of Protein-Protein Interactions

Protein-protein interactions (PPIs) are critical for organizing molecules in a cell and mediating signaling pathways. Dysregulation of PPIs is often a key driver of disease. To better understand the biophysical basis of such disease processes-and to potentially target them-it is critical to understand the molecular determinants of PPIs. Deep mutational scanning (DMS) facilitates the acquisition of large amounts of biochemical data by coupling selection with high throughput sequencing (HTS). The challenging and labor-intensive design and optimization of a relevant selection platform for DMS, however, limits the use of powerful directed evolution and selection approaches. To address this limitation, we designed a versatile new phage-assisted continuous selection (PACS) system using our previously reported proximity-dependent split RNA polymerase (RNAP) biosensors, with the aim of greatly simplifying and streamlining the design of a new selection platform for PPIs. After characterization and validation using the model KRAS/RAF PPI, we generated a library of RAF variants and subjected them to PACS and DMS. Our HTS data revealed positions along the binding interface that are both tolerant and intolerant to mutations, as well as which substitutions are tolerated at each position. Critically, the "functional scores" obtained from enrichment data through continuous selection for individual variants correlated with K_D values measured in vitro, indicating that biochemical data can be extrapolated from sequencing using our new system. Due to the plug and play nature of RNAP biosensors, this method can likely be extended to a variety of other PPIs. More broadly, this, and other methods under development support the continued development of evolutionary and high-throughput approaches to address biochemical problems, moving toward a more comprehensive understanding of sequence-function relationships in proteins.

iSeq 2.0: A Modular and Interchangeable Toolkit for Interaction Screening in Yeast

We developed a flexible toolkit for combinatorial screening in Saccharomyces cerevisiae, which generates large libraries of cells, each uniquely barcoded to mark a combination of DNA elements. This interaction sequencing platform (iSeq 2.0) includes genomic landing pads that assemble combinations through sequential integration of plasmids or yeast mating, 15 barcoded plasmid libraries containing split selectable markers (URA3AI, KanMXAI, HphMXAI, and NatMXAI), and an array of ∼24,000 "double-barcoder" strains that can make existing yeast libraries iSeq compatible. Various DNA elements are compatible with iSeq: DNA introduced on integrating plasmids, engineered genomic modifications, or entire genetic backgrounds. DNA element libraries are modular and interchangeable, and any two libraries can be combined, making iSeq capable of performing many new combinatorial screens by short-read sequencing.

Selection of Protein-Protein Interactions of Desired Affinities with a Bandpass Circuit

We have developed a genetic circuit in Escherichia coli that can be used to select for protein-protein interactions of different strengths by changing antibiotic concentrations in the media. The genetic circuit links protein-protein interaction strength to β-lactamase activity while simultaneously imposing tuneable positive and negative selection pressure for β-lactamase activity. Cells only survive if they express interacting proteins with affinities that fall within set high- and low-pass thresholds; i.e. the circuit therefore acts as a bandpass filter for protein-protein interactions. We show that the circuit can be used to recover protein-protein interactions of desired affinity from a mixed population with a range of affinities. The circuit can also be used to select for inhibitors of protein-protein interactions of defined strength.

Functional Annotation of Bacterial Signal Transduction Systems: Progress and Challenges

Bacteria possess a large number of signal transduction systems that sense and respond to different environmental cues. Most frequently these are transcriptional regulators, two-component systems and chemosensory pathways. A major bottleneck in the field of signal transduction is the lack of information on signal molecules that modulate the activity of the large majority of these systems. We review here the progress made in the functional annotation of sensor proteins using high-throughput ligand screening approaches of purified sensor proteins or individual ligand binding domains. In these assays, the alteration in protein thermal stability following ligand binding is monitored using Differential Scanning Fluorimetry. We illustrate on several examples how the identification of the sensor protein ligand has facilitated the elucidation of the molecular mechanism of the regulatory process. We will also discuss the use of virtual ligand screening approaches to identify sensor protein ligands. Both approaches have been successfully applied to functionally annotate a significant number of bacterial sensor proteins but can also be used to study proteins from other kingdoms. The major challenge consists in the study of sensor proteins that do not recognize signal molecules directly, but that are activated by signal molecule-loaded binding proteins.

What We Do Not Know about Fungal Cell Adhesion Molecules

There has been extensive research on structure and function of fungal cell adhesion molecules, but the most of the work has been about adhesins in Candida albicans and Saccharomyces cerevisiae. These yeasts are members of a single ascomycete order, and adhesion molecules from the six other fungal phyla are only sparsely described in the literature. In these other phyla, most of the research is at the cellular level, rather than at the molecular level, so there has been little characterization of the adhesion molecules themselves. A catalog of known adhesins shows some common features: high Ser/Thr content, tandem repeats, N- and O-glycosylations, GPI anchors, dibasic sequence motifs, and potential amyloid-forming sequences. However, none of these features is universal. Known ligands include proteins and glycans on homologous cells and host cells. Existing and novel tools can exploit the availability of genome sequences to identify and characterize new fungal adhesins. These include bioinformatics tools and well-established yeast surface display models, which could be coupled with an adhesion substrate array. Thus, new knowledge could be exploited to answer key questions in fungal ecology, animal and plant pathogenesis, and roles of biofilms in infection and biomass turnover.