Shotgun scanning glycomutagenesis: A simple and efficient strategy for constructing and characterizing neoglycoproteins

Carbohydrate-active enzyme (CAZyme) discovery and engineering via (Ultra)high-throughput screening

Carbohydrate-active enzymes (CAZymes) constitute a diverse set of enzymes that catalyze the assembly, degradation, and modification of carbohydrates. These enzymes have been fashioned into potent, selective catalysts by millennia of evolution, and yet are also highly adaptable and readily evolved in the laboratory. To identify and engineer CAZymes for different purposes, (ultra)high-throughput screening campaigns have been frequently utilized with great success. This review provides an overview of the different approaches taken in screening for CAZymes and how mechanistic understandings of CAZymes can enable new approaches to screening. Within, we also cover how cutting-edge techniques such as microfluidics, advances in computational approaches and synthetic biology, as well as novel assay designs are leading the field towards more informative and effective screening approaches.

Structure-based neural network protein-carbohydrate interaction predictions at the residue level

Carbohydrates dynamically and transiently interact with proteins for cell-cell recognition, cellular differentiation, immune response, and many other cellular processes. Despite the molecular importance of these interactions, there are currently few reliable computational tools to predict potential carbohydrate-binding sites on any given protein. Here, we present two deep learning (DL) models named CArbohydrate-Protein interaction Site IdentiFier (CAPSIF) that predicts non-covalent carbohydrate-binding sites on proteins: (1) a 3D-UNet voxel-based neural network model (CAPSIF:V) and (2) an equivariant graph neural network model (CAPSIF:G). While both models outperform previous surrogate methods used for carbohydrate-binding site prediction, CAPSIF:V performs better than CAPSIF:G, achieving test Dice scores of 0.597 and 0.543 and test set Matthews correlation coefficients (MCCs) of 0.599 and 0.538, respectively. We further tested CAPSIF:V on AlphaFold2-predicted protein structures. CAPSIF:V performed equivalently on both experimentally determined structures and AlphaFold2-predicted structures. Finally, we demonstrate how CAPSIF models can be used in conjunction with local glycan-docking protocols, such as GlycanDock, to predict bound protein-carbohydrate structures.

Rapid biosynthesis of glycoprotein therapeutics and vaccines from freeze-dried bacterial cell lysates

The advent of distributed biomanufacturing platforms promises to increase agility in biologic production and expand access by reducing reliance on refrigerated supply chains. However, such platforms are not capable of robustly producing glycoproteins, which represent the majority of biologics approved or in development. To address this limitation, we developed cell-free technologies that enable rapid, modular production of glycoprotein therapeutics and vaccines from freeze-dried Escherichia coli cell lysates. Here, we describe a protocol for generation of cell-free lysates and freeze-dried reactions for on-demand synthesis of desired glycoproteins. The protocol includes construction and culture of the bacterial chassis strain, cell-free lysate production, assembly of freeze-dried reactions, cell-free glycoprotein synthesis, and glycoprotein characterization, all of which can be completed in one week or less. We anticipate that cell-free technologies, along with this comprehensive user manual, will help accelerate development and distribution of glycoprotein therapeutics and vaccines.

Profiling Germinal Center-like B Cell Responses to Conjugate Vaccines Using Synthetic Immune Organoids

Glycoengineered bacteria have emerged as a cost-effective platform for rapid and controllable biosynthesis of designer conjugate vaccines. However, little is known about the engagement of such conjugates with naïve B cells to induce the formation of germinal centers (GC), a subanatomical microenvironment that converts naïve B cells into antibody-secreting plasma cells. Using a three-dimensional biomaterials-based B-cell follicular organoid system, we demonstrate that conjugates triggered robust expression of hallmark GC markers, B cell receptor clustering, intracellular signaling, and somatic hypermutation. These responses depended on the relative immunogenicity of the conjugate and correlated with the humoral response in vivo. The occurrence of these mechanisms was exploited for the discovery of high-affinity antibodies against components of the conjugate on a time scale that was significantly shorter than for typical animal immunization-based workflows. Collectively, these findings highlight the potential of synthetic organoids for rapidly predicting conjugate vaccine efficacy as well as expediting antigen-specific antibody discovery.

Structure-Based Neural Network Protein-Carbohydrate Interaction Predictions at the Residue Level

Carbohydrates dynamically and transiently interact with proteins for cell-cell recognition, cellular differentiation, immune response, and many other cellular processes. Despite the molecular importance of these interactions, there are currently few reliable computational tools to predict potential carbohydrate binding sites on any given protein. Here, we present two deep learning models named CArbohydrate-Protein interaction Site IdentiFier (CAPSIF) that predict carbohydrate binding sites on proteins: (1) a 3D-UNet voxel-based neural network model (CAPSIF:V) and (2) an equivariant graph neural network model (CAPSIF:G). While both models outperform previous surrogate methods used for carbohydrate binding site prediction, CAPSIF:V performs better than CAPSIF:G, achieving test Dice scores of 0.597 and 0.543 and test set Matthews correlation coefficients (MCCs) of 0.599 and 0.538, respectively. We further tested CAPSIF:V on AlphaFold2-predicted protein structures. CAPSIF:V performed equivalently on both experimentally determined structures and AlphaFold2 predicted structures. Finally, we demonstrate how CAPSIF models can be used in conjunction with local glycan-docking protocols, such as GlycanDock, to predict bound protein-carbohydrate structures.

Ribosome Stalling of N-Linked Glycoproteins in Cell-Free Extracts

Ribosome display is a powerful in vitro method for selection and directed evolution of proteins expressed from combinatorial libraries. However, the ability to display proteins with complex post-translational modifications such as glycosylation is limited. To address this gap, we developed a set of complementary methods for producing stalled ribosome complexes that displayed asparagine-linked (N-linked) glycoproteins in conformations amenable to downstream functional and glycostructural interrogation. The ability to generate glycosylated ribosome-nascent chain (glycoRNC) complexes was enabled by integrating SecM-mediated translation arrest with methods for cell-free N-glycoprotein synthesis. This integration enabled a first-in-kind method for ribosome stalling of target proteins modified efficiently and site-specifically with different N-glycan structures. Moreover, the observation that encoding mRNAs remained stably attached to ribosomes provides evidence of a genotype-glycophenotype link between an arrested glycoprotein and its RNA message. We anticipate that our method will enable selection and evolution of N-glycoproteins with advantageous biological and biophysical properties.

Strategies for Glycoengineering Therapeutic Proteins

Almost all therapeutic proteins are glycosylated, with the carbohydrate component playing a long-established, substantial role in the safety and pharmacokinetic properties of this dominant category of drugs. In the past few years and moving forward, glycosylation is increasingly being implicated in the pharmacodynamics and therapeutic efficacy of therapeutic proteins. This article provides illustrative examples of drugs that have already been improved through glycoengineering including cytokines exemplified by erythropoietin (EPO), enzymes (ectonucleotide pyrophosphatase 1, ENPP1), and IgG antibodies (e.g., afucosylated Gazyva®, Poteligeo®, Fasenra™, and Uplizna®). In the future, the deliberate modification of therapeutic protein glycosylation will become more prevalent as glycoengineering strategies, including sophisticated computer-aided tools for "building in" glycans sites, acceptance of a broad range of production systems with various glycosylation capabilities, and supplementation methods for introducing non-natural metabolites into glycosylation pathways further develop and become more accessible.