Glycoengineering is the manipulation of protein glycosylation patterns with the goal of controlling or limiting heterogeneity and improving predictability. The glycoengineer's approach to glycobiology thus turns a descriptive field into a prescriptive one, where sugars are viewed not as the result of "garbling" (Rutter) but as modular codes whose values must be programmed. Glycoengineering involves a broad range of tactics to reprogram host cells for the purpose of manipulating the composition, topology, and site occupancy of glycans. The result of successful glycoengineering is the ability to convert the diverse mixture of glycoforms typically generated by cell culture into homogeneous glycoforms of desired biological attributes. Glycoengineering has evolved from the early days of simple glycosyltransferase overexpression to include sophisticated systems for simultaneous CRISPR knockout of undesirable biosynthetic pathways, redirection of metabolic flux with small molecules, and chemoenzymatic remodeling of purified protein. The use of glycoengineering in this manner is particularly relevant to biologics, as glycosylation is now recognized by regulatory authorities to be a critical quality attribute (CQA) of biopharmaceuticals. As such, glycoengineering is an important tool in the development of a wide range of biologics, including monoclonal antibodies, replacement enzymes, and synthetic vaccines. In addition to the classic "wet lab" approaches in living cells, a host of new glycoengineering technologies have been developed in recent years that would not require sophisticated infrastructure or technical know-how, and thus has the potential to be widely disseminated outside of glycoengineering labs. Among these are light-activated glycosyltransferases and cell-free biosynthetic cascades.
Glycoengineering is an outgrowth of the realization that despite having undergone optimization through evolutionary pressures, natural cellular glycosylation processes are not well matched for the reproducible manufacture of biotherapeutics. Glycoengineering is a discipline that merges the fundamental carbohydrate chemistry underpinning with process engineering design space thinking and manufacturing know-how, resulting in a community that views glycan structures as design primitives, modifiable interchangeable building blocks with free reagent access, rather than fixed, unalterable, and recalcitrant artifacts of cellular metabolism. The degree of heterogeneity that had long been accepted as unavoidable is instead replaced by glycoengineering decision points wherein oligosaccharides are viewed as unique design handles that can be added, subtracted, or selectively masked toward meeting therapeutic needs. Glycoengineering concepts and tactics have led to the development of host-cell lines in which fucosyltransferases are silenced, fed-batch formulations in which metabolic building blocks are supplemented and terminally added enzymes that "polish" glycans after harvest but before formulation, all delivering on the promise of defined glyco-codes. In addition to impacting clone selection decisions, process considerations now span upstream fermentation as well as downstream purification and have impacted analytical release decisions. Glycoengineering is the enabling technology that has allowed the conversion of mechanistic glycobiology to the manufacture of medicines with predictable clinical performance.
Glycoengineering is the strategic, rational design and modification of glycans covalently attached to therapeutic proteins to enable specific drug development goals, rather than non-specific glycosylation as a consequence of the host metabolism. Glycoengineering is based on the concept of carbohydrates as "informational" or "addressable" components where their composition, branching and terminal groups precisely tune protein folding and stability, receptor interactions and immunogenicity. Glycoengineering is important because therapeutic activity may be dependent on glycan epitopes in addition to the amino acid sequence for pharmacokinetics and effector functions. Techniques based on metabolic engineering (feeding cells with fluorinated fucose analogues to inhibit alternative pathways), as well as genetic engineering to knock down glycosyltransferases can be used to ensure that glycoform distributions fall within desired specification ranges (such as tight regulatory ranges for product approval). In addition to monoclonal antibodies, glycoengineering is applied to protein therapeutics like enzyme replacement therapies where mannose-6-phosphate is engineered to enable lysosomal targeting and synthetic vaccines where tumor associated carbohydrate antigens are presented on protein scaffolds. Methods like chemoenzymatic remodeling have also been developed to add non-natural sugars after protein production, to further expand the chemical space beyond what is natively possible. Glycoengineering aims to create a more predictable glycosylation that reduces glycans as a manufacturing variable and enables glycoproteins to be customized to a patient's glyco-profile. As clinical data is generated to show specific glycoforms correlate with patient response, glycoengineering will become increasingly important for the development of biosimilars with equivalent glycophenotypes and next generation biologics with improved safety and efficacy profiles.
Differences between natural glycosylation and glycoengineering are best highlighted when compared as a point of departure: naturally, glycosylation is stochastic (chemically random) in nature. Glycosylation at the monomer level is determined by the relative presence of individual enzymes, the availability of sugar nucleotide pools and transit times through the Golgi system. The glycoform mixture at the protein level results in a heterogeneously glycosylated population where each protein molecule has a different glycan. The heterogeneity will also vary depending on the cell-line used to produce the protein. For example, CHO cells display fucosylated but non-galactosylated patterns whereas yeast will elaborate high-mannose structures which are not therapeutic in humans. The result is also generally undesigned, by-products of protein production and glycan assembly that have been naturally selected for, such that their evolutionary purpose optimises cellular fitness (rather than pharmaceutical utility). In comparison, glycoengineering is by definition a deterministic (predictable) process where glycosylation is designed and each step is controlled, through rational design. This can be achieved by silencing competing glycosylation pathways, supplementing precursor availability or remodelling heterogenous glycan pools after production. Glycoengineering is also not limited to the monosaccharide pool present in a host cell and can be expanded to include non-natural sugars, synthetic linkages or chemical tags that can be used to increase protein functionality. Predictability is also important, with naturally occurring glycosylation being very sensitive to changes in the overall production process, while engineered systems can be built to be stable and process agnostic. The intended result is also distinct from glycosylation where the glycan is a by-product of protein trafficking and folding. Glycoengineering is in place for a purpose, most often to tune a glycoprotein for human therapeutic needs such as half-life extension, tissue targeting and/or immune modulation. This re-characterization of carbohydrates shifts the perception of carbohydrates from being just post-translational modifications, into 'active' pharmaceutical ingredients, and as such, glycoengineering should be considered a discipline in its own right, applying the principles of engineering to biology's complexity.
A schematic of different glyco-engineering strategies; (A) protein engineering and (B) cell engineering.1,5
The strategies of glycoengineering are, essentially, based on the ability to rationally rewire cellular machinery to replace the stochastic nature of cellular pathways with a directed process. Glycoengineering is based on the general assumption that the glycosylation process is malleable and can be seen as an assembly line that can be subverted into a desired one. The process starts with a cell line choice: although CHO cells are the mainstay of production due to the availability of cell lines and human-like glycan expression, their glycosylation is highly heterogeneous and they need to be engineered to remove undesirable and potentially immunogenic sugar epitopes, such as α-1,3-fucose or NGNA sialic acids. The second key principle in glycoengineering is to modulate the metabolic flux towards the various glycosylation pathways by blocking or adding alternative unnatural sugars, analogues, or precursors. A major strength of current glycoengineering practices is the availability of methods with strong genetic precision, such as CRISPR-based knockouts, knockdowns using siRNA, or transgenic overexpression, to irreversibly redirect the biosynthetic machinery. A third strategy in glycoengineering consists of remodeling glycan structures after biosynthesis is complete. This approach makes use of endoglycosidases to reduce the glycans to a common homogeneous structure to which any desired sugar can be attached using purified glycosyltransferases in the presence of activated sugar donors. In general, the key principles of glycoengineering aim to render the glycosylation process more predictable, and thus controllable, with the goal of manufacturing a glycoform within a defined specification range, rather than a product that is subject to biological variability.
Site-specific glycoengineering necessitates prior knowledge of consensus sequences. For N-glycosylation, the canonical N-glycosylation sequon Asn-X-Ser/Thr is required but not sufficient and predicted by computational tools like NetNGlyc, with validation necessary on a case-by-case basis. Rational site selection for neoglycans can be determined based on known structures or homology models of a target protein that feature large surface-exposed loops that may accommodate engineered glycans without interfering with protein folding or occluding active sites. Site-saturation mutagenesis, the randomization of each codon at positions 2, 3 and 4 of the target site has been used in combination with screening for functional proteins that accept the novel glycan. An important caveat is steric shielding, where attached glycans may block the binding of a ligand or contribute to the formation of a patch on the surface of a protein that facilitates aggregation. In contrast, they can also serve to protect labile regions of a protein from proteolysis, as has been shown for ENPP1-Fc, in which built-in glycans shielded the protein from serum proteases. Site selection for O-glycosylation, which lacks a consensus sequon, is achieved by random mutagenesis of serine/threonine-rich regions and necessitates empirical optimization with glycosyltransferases. Success in engineered glycan attachment and occupancy is usually judged by whether it confers similar or improved pharmacokinetics to that of endogenous sugars, without creating an immunogenic structure.
There are three main genetic, metabolic, and enzymatic tools for glycoengineering that can be used in series or in parallel. Genome editing using CRISPR/Cas9 nucleases makes it possible to knockout glycosyltransferases such as Fut8 or to knock in humanized versions of glycogenes into a safe harbor location for stable expression. Zinc-finger nucleases and TALENs have also been developed for gene disruption, but CRISPR is the more widely used system because of the ability to multiplex this technology. Metabolic engineering strategies are also used to block or install sugars in proteins, such as the competitive inhibition of fucosylation through the addition of fluorinated sugar analogues to cells or the addition of azido-sugars that can be chemically conjugated to proteins in a bioorthogonal manner without genetic manipulation. The structure of glycans attached to proteins can also be remodelled in-vitro after production, using chemoenzymatic methods. Endoglycosidases can be used to trim down heterogeneous glycans to a consistent GlcNAc stub, for example, which can then be rebuilt with the correct structure using purified glycosyltransferases and sugar nucleotides, including those with unnatural sugars. Scaffold-mediated enzyme methods also spatially organize several enzymes with DNA-directed self-assembly in a manner that can mimic the organization in the Golgi and increase reaction efficiency. Sequential immobilization of glycosyltransferases onto resins also allows for their reuse, which is important for these often expensive enzymes that would normally be used and discarded once. Orthogonal methods such as high-resolution mass spectrometry or HILIC-FLD can be used for quality control to confirm that the glycoforms produced match the desired engineered glycan structure.
A variety of tools are being developed to gain more control over glycosylation processes, including synthetic biology, genetic engineering and enzyme engineering. Synthetic biology tools allow the glycosylation processes of an organism to be more completely understood and manipulated. Genome editing enables multi-gene knockout to prevent unwanted competition from native enzymes for substrates, so that glycosylation can be engineered with a single type of glycoform, known as homogeneity. Metabolic engineering allows cells to be supplemented with chemically modified sugar substrates that contain a bioorthogonal functional group. For example, these substrates are bioorthogonal in the sense that they are otherwise natural monosaccharides. These modified sugars are metabolized by the cell and incorporated into glycoproteins in place of the natural sugars. By using modified sugars with chemically tagged groups, more complex sugars can be produced. In addition, cells can be engineered to use non-natural sugars. This process can also be combined with techniques in protein engineering to create glycoproteins with a consistent, uniform glycan structure. Solid-phase enzymatic remodeling has also allowed the introduction of more complex, non-natural glycan structures after production. In this technique, the glycosyltransferase enzyme is immobilized on a resin, and the protein is sequentially passed over it through repeated reaction steps, which do not require purification between steps. Remodeling using this method has been shown to not affect protein folding. In addition, transient expression in plants offers an opportunity for the on-demand, regionalized production of glycoproteins. In this method, genes necessary for glycosylation of a protein of interest are delivered into a plant using a viral vector. In just a few days, the plant glycome can be humanized to glycoengineer the desired glycoprotein. Synthetic glycobiology also now allows the creation of artificial pathways of glycosylation in bacteria, or in cell-free systems, thus bypassing mammalian cells completely. In addition, the emerging use of machine learning in the field of glycomics to analyze data and make predictions could reduce the need for trial-and-error experimentation.
The development of CRISPR/Cas9-based genome editing has accelerated the adoption of glycoengineering in the industry. It is now possible to "rewrite" the glycosylation potential of a mammalian production cell line at high speed and low cost with previously unmatched precision. In contrast to previous zinc-finger nuclease or TALEN-based strategies that required the development of a new protein construct for each new target gene, CRISPR/Cas9 allows the co-transfection of several guide RNAs to inactivate multiple glycosyltransferases at once (multiplexing). The most widely used application of this method is to silence FUT8, the only enzyme that mediates core fucosylation, to produce afucosylated antibodies with improved ADCC potency. The knock-in of human glycosyltransferases using CRISPR/Cas9 into non-mammalian expression hosts has also been demonstrated. Transgenic expression of GnT-I and GnT-II in yeast cells converts high-mannose glycans to complex-type patterns, humanizing the glycome of these microbial cell factories. CRISPR activation (CRISPRa) of limiting enzymes such as GnT-III to bisect glycans and CRISPR interference (CRISPRi) to repress other competing glycosylation pathways are being developed to upregulate or downregulate glycosylation pathways without making permanent changes to the genome, potentially allowing the fine-tuning of glycosylation in cell lines used in process development. CRISPR also permits allele-specific editing of duplicate glycosyltransferase paralogs, which often cannot be knocked out using traditional strategies. In plant cell glycoproteins, CRISPR can be used to remove undesired immunogenic β1,2-xylose and α1,3-fucose residues, which would otherwise produce anti-carbohydrate antibodies in humans upon therapeutic use. Although off-target effects are a possibility with CRISPR, high-fidelity Cas9 variants and thorough sequencing validation of the final cell lines are used to ensure no unwanted mutations are present and the modified cells remain stable in culture over many passages. CRISPR allows screening of hundreds of clones in a short time, dramatically accelerating cell-line development programs.
Synthetic biologists have created artificial glycosyltransferases by rationally engineering or evolving enzyme catalytic domains to have different specificities and improved biocompatibility. Glycosynthases are mutants of glycosyltransferases engineered to have no hydrolytic activity, only to use the donor substrate to glycosylate a protein at the acceptor site. These engineered enzymes can be used to install linkages that do not naturally occur because there is no hydrolysis in the reverse direction. Random mutagenesis followed by screening has also allowed for substrate expansion to larger bulky aglycones or sugars with non-natural chemical groups like fluorine that the wild-type enzyme would not have tolerated. Chemoenzymatic protein remodeling platforms in solid phase have been developed, which is advantageous because the glycoprotein substrate can be immobilized on a resin and multiple enzymes can be added one after another to catalyze a series of remodeling steps without the need for buffer exchange, saving time and optimizing yields of each step. Light-activated enzymes have also been engineered by fusing a photodimerizable domain to the enzyme, inactivating it until it is activated with UV light. This method allows glycosylation to be temporally controlled, and only added after protein folding is completed. Cell-free extracts that combine purified enzymes and sugar nucleotides are also advantageous as glycoengineering of this sort would be toxic to cells if expressed in vivo. The reactions are also scalable to microfluidic droplet reactors that allow these reactions to be performed in droplets on picogram amounts of protein. As the cost of these enzymes decrease by immobilization or directed evolution, these methods have transitioned from proof of principle to GMP-compatible processes for the in vitro production of improved biotherapeutics. One application of these processes has been post-translational remodeling, such as the conversion of off-the-shelf biologics into their improved forms without starting from cell-line development. One such example is treating an approved antibody in vitro with an α-2,6-sialyltransferase and CMP-sialic acid to form anti-inflammatory glycoforms.
A number of biologics and biopharmaceuticals have been glycoengineered with improved or novel characteristics compared to their native counterparts. For example, glycoengineering has been used to improve the pharmacokinetic properties and modulate immunogenicity and immune-cell interactions in a wide range of monoclonal antibodies for various types of cancer and autoimmune diseases, as well as to design new synthetic vaccines. Glycoengineering can also be used to fine-tune biologic glycosylation to improve clinical performance, increase efficacy and safety, and is being implemented as a quality attribute to engineer more consistent biologics from batch to batch and between biologic products from different suppliers. Pharmaceutical companies are harnessing this emerging technology and incorporating glycan design considerations early in the drug development process, and regulatory agencies are requiring more detailed glycan analysis and characterization to ensure product consistency. Glycoengineering can be applied to already existing biologic modalities or to create new biologics that could not otherwise be created, such as asymmetric bispecific antibodies and multi-valent glycoconjugate vaccines. In addition to the development of new human therapeutics, glycoengineering has been applied to animal and veterinary therapeutics in order to optimize glycosylation for target animal species, as well as to industrial enzymes for improved stability in industrial conditions. Glycoengineering is also being used in diagnostics. Engineered glycan-binding proteins and glycan arrays can be used as orthogonal approaches to detect glycan biomarkers. The glycoengineering field is expected to continue to rapidly expand and advance with new tools and technologies, such as computational modeling of glycosylation and structure/function relationships, synthetic biology, and high-throughput glycomics and glycoproteomics analysis methods. In addition, development of cell-free glycosylation platforms that would allow for decentralization of biologic production and on-demand rapid production in response to pandemics and other global needs is also advancing rapidly.
Glycoengineering is a major approach applied in the biopharmaceutical industry to improve therapeutic monoclonal antibodies (mAbs) and other recombinant proteins. The first commercial application was the use of glycoengineering to improve ADCC of mAbs. The lack of core fucose in the Fc glycans due to genetic knockout or inhibition of metabolic pathways results in higher affinity for the activating Fc-γRIIIa receptor on NK cells and better tumor cell killing without higher doses or increased toxicity, as shown by approved drugs for hematological cancers. Other key applications in biopharmaceutical glycoengineering include: reduction of immunogenicity by expressing humanized glycans (without α-1,3 galactose, NGNA sialic acid, etc.), improvement of pharmacokinetics (PK) by increasing sialylation to mask galactose (terminal sugar) and avoid recognition by asialoglycoprotein receptor (ASGPR) on hepatocytes, improved consistency for biosimilars (mAbs in this case) by matching the glycan profile with the reference product, better delivery in enzyme replacement therapies (ERTs) by increasing mannose-6-phosphate (M6P) groups on the N-glycans to achieve lysosomal targeting, and adjusting tissue penetration for fusion proteins by tuning the size of glycans. These and more recent modalities, such as bispecific antibodies and Fc-fusion proteins, also rely on glycoengineering approaches to improve and control other functions at the same time. Common strategies to engineer glycans on therapeutic proteins include genetic engineering of the glycosylation machinery in host cells, metabolic engineering of host cells by supplementing precursors (metabolites) in the glycan biosynthetic pathway to shift glycan profile, and post-purification enzymatic remodeling of glycans on therapeutic proteins to remove and add glycans. This toolbox of approaches allows the protein developers to control glycan heterogeneity that is introduced at different stages of the process to ensure therapeutic proteins are delivered with well-defined, homogenous glycoforms for predictable clinical results.
Glycoengineering is applied for the production of glycoconjugate vaccines to prevent invasive disease caused by bacteria. In this application, polysaccharides antigens are purified from the pathogen and covalently conjugated to protein carriers to boost the immune response. The innate response to the carbohydrate antigen is normally T-independent but conjugation to a protein carrier results in T-dependent immunity, producing a long-lived protective response. Glycoengineering can be used in this context to improve on the standard procedure; well-defined, homogeneous glycan antigens eliminate batch-to-batch variation in polysaccharide antigens extracted from natural sources and can therefore be more rapidly developed through regulatory approval and manufacturing scale-up. Glycoengineering is also used to generate synthetic carbohydrate antigens for therapeutic cancer vaccination. In these vaccines, tumor-associated carbohydrate epitopes such as GM3 ganglioside, Globo H hexasaccharide, and Lewis Y antigens are conjugated to a carrier protein to produce antibodies against malignant cells. Synthetic antigens can be readily modified to identify which epitopes are immunodominant through glycan microarray screening and facilitate rational antigen selection. N-acetylgalactosamine (GalNAc) residues can be targeted to antigen presenting cells through C-type lectin receptors to increase uptake and drive mucosal immunity, an important goal for pathogens entering the body via the respiratory and reproductive tracts. The recent development of mRNA vaccines has introduced new glycoengineering challenges for vaccine development, as the antigen must be properly folded and glycosylated upon in-vitro translation to mimic native viral glycoproteins. The choice of host cell and metabolic engineering to balance cell growth with glycosylation is important for vaccine glycoproteins. Glycoengineering of bacterial systems are well-suited to respond to emerging pathogens because the glycoenzymes are easily purified for a cell-free system in which the lipid-linked oligosaccharide is assembled on a purified protein scaffold. Glycoprotein synthesis can be performed in cell-free systems using bacterial extracts from glyco-competent E. coli strains. Cell-free synthesis of glycoproteins has been adapted for the production of conjugate vaccines, and eliminates the need for living cells, reducing purification complexity and potential endotoxin contaminants.
Fig. 2 Glycoengineering approaches to produce authentic viral N-glycan (A) and O-glycan (B) structures in plants.2,5
Critical issues: Glycoengineering is highly complex. There are a number of factors that contribute to this, including: Glycans are synthesized by a multi-step enzymatic process. Each step is catalyzed by a different glycosyltransferase and requires the use of a specific sugar nucleotide donor substrate. This lack of control can be magnified, as the first step in the pathway can have downstream effects on the rest of the glycan processing by the Golgi. This may result in unintended glycoforms that make the glycoprotein product a mixture. Glycosyltransferases used for glycoengineering are often purified for a single use. The reagents are typically expensive and lack stability, which makes scaling of the process more difficult. While immobilization of enzymes onto a resin is an option, it needs to be proven that there is no leaching of the enzyme and that there is no loss of activity over multiple cycles. Production robustness can be an issue as even minor changes in pH, dissolved oxygen, or metal ions in the culture can alter the metabolism, resulting in different glycan profiles. This would result in glycans outside of the specification range and ultimately cause regulatory issues. Analytical methods, such as HPLC, MS, and HILIC-FLD, require additional training and investment and can become a bottleneck for release testing. Substrate specificities for a number of glycosyltransferases have yet to be determined. O-glycosylation in particular is an issue as the polypeptide GalNAc-transferases, which act as initiators of O-glycosylation, are not well characterized. As for glycosylation processing by the Golgi, there are many factors that are not understood such as what dictates Golgi residence time, transport mechanisms, and other factors which make processing outcomes difficult to predict.
Opportunities: Machine learning and artificial intelligence can be used to identify patterns in glycomics data to predict structure/function. CRISPR technologies can be used for multiplexing to make alterations in pathways. Cell-free systems can be used to avoid cellular toxicity issues, making glycoengineering an attractive tool for biomanufacturing.
Current limitations in glycoengineering technology are inherently non-linear and extremely varied. The stepwise requirements of multiple glycosyltransferases and nucleotide sugar donors in the Golgi apparatus result in kinetic limitations, with earlier enzymes being rate-limiting (precursor trapping and shunting of intermediate glycoforms) and later enzymes in a substrate limitation (synthesis of truncated and potentially therapeutically inadequate forms). Furthermore, the inability to reuse glycosyltransferases is also a limitation. Glycosyltransferases are unstable proteins and are typically used once before losing activity, making in-vitro remodeling very expensive, despite its potential for exactness and control. Immobilizing glycosyltransferases to a resin allows reuse, but there is the added difficulty of correctly orienting enzymes and maintaining enzyme integrity during repeated cycles of binding and elution, as well as ensuring that leaked pieces of enzyme do not co-purify with the target protein. Glycoengineering efforts can also be compromised by variations in cell line production. In some cases, even clonal populations of cells will have small changes in their metabolism over a long culture period and can alter the glycosylation profile as a result. Other environmental factors, such as small temperature differences or a new lot of media, can alter the expression of glycosyltransferases enough to result in glycan profiles that fall outside of validated limits. Analytical tools are also extremely useful, but require a high degree of expertise to be used properly. Mass spectrometric analysis of isomeric glycans is often inconclusive without enzymatic digestion, and HILIC quantitation is dependent on carefully controlled mobile phase composition and column temperature, complicating method transfer between labs. Chemoenzymatic glycosylation is also limited in its ability to introduce non-natural sugars or topologies, as most glycosyltransferases are very specific for both their donor and acceptor substrates. Recombinant protein engineering must often be used to gain more control over glycosylation. Lastly, multi-modal glycoengineering (i.e. both N- and O-glycosylation) is still a work in progress. These two glycosylation pathways compete for nucleotide sugar donors and residence time in the Golgi and, as a result, attempts to modify both at the same time often result in unexpected results.
Synthetic biology approaches, systems and computational modelling, and innovative manufacturing concepts will all influence the next generation of glycoengineering. Machine learning could make its mark by helping to identify optimal glyco-engineering approaches based on sequence and structural databases without extensive trial-and-error. Algorithms may also be used to minimise empirical screening and recognise non-obvious enzyme cocktails that co-express to create specific glycoforms. CRISPR systems will mature past basic gene deletion to become programmable expression tools, where glycosyltransferases can be regulated by artificial inducible promoters to add layers of temporal control over glycan elaboration during fermentation or dynamically adjust in response to changing metabolic states. Cell-free glycoprotein synthesis will disrupt established bioprocessing workflows by fusing purified glycosyltransferases with artificial nucleotide sugars in microfluidic systems to produce glycoforms not possible or practical inside living cells, bypassing regulatory concerns over recombinant organisms. Directed evolution of glycosyltransferases will increase their substrate range and allow the creation of artificial enzymes that accept bulky aglycones, fluorinated sugars, or click-chemistry linkers for the installation of non-native functionalities amenable to site-specific payload conjugation. Glycomics will be high-throughput by automating sample prep, HILIC-FLD analysis, and data analysis workflows, reducing the analytical timeline from weeks to days and bringing real-time glycan monitoring to continuous manufacturing. Standardization efforts will converge on open-source glycan structural databases and validated informatics workflows that automatically annotate glycomics data to improve cross-laboratory reproducibility. Interdisciplinary crossovers with nanotechnology may lead to glycan-functionalized nanoparticles for targeted delivery, or plant-based transient expression systems could level the playing field for low-resource vaccine manufacturing. In the end, these advances will likely bring glycoengineering out of the realm of boutique production and into the mainstream as a modular, scalable platform integrated into biomanufacturing workflows to rapidly prototype personalized glycoproteins with sugar signatures matched to individual patient glycoprofiles and disease mechanisms.
Originally a somewhat obscure technical art, glycoengineering now stands as a foundational practice in modern biopharmaceutical development. Glycoengineering is characterized by the reversal of the stochastic nature of host-cell glycosylation and its conversion into a deterministic and programmable system. This approach allows for the manipulation of protein glycosylation to produce well-defined glycoproteins with specified carbohydrate structures that are no longer random or unpredictable but rather are intentionally designed to modulate therapeutic efficacy. This is possible through the convergence of three core enabling technologies: CRISPR-based host engineering, cell-free glycosynthesis, and machine learning-based glyco-design. Each of these advances mitigates the three major pre-existing challenges facing the glycoengineering space: enzyme instability, metabolic pathway flux, and analytics bottlenecks. Glycoengineering has been recognized by regulatory authorities such as the FDA as well as the biopharmaceutical industry, where the glycan distribution has been made a key critical quality attribute that must be defined and qualified.
Unlock precise control over glycosylation with our advanced glycoengineering and custom glycan modification services. Using state-of-the-art enzymatic, chemoenzymatic, and genetic engineering technologies, we help you design and modify glycan structures to enhance protein performance, improve therapeutic efficacy, and meet biopharmaceutical quality standards. Our team integrates high-resolution analytics—LC-MS/MS, HILIC-HPLC, CE-MS, and glycan mapping—to ensure every engineered modification is accurately characterized and functionally validated. Our glycoengineering solutions empower you to:
Whether you're innovating new therapeutic modalities, refining biologic quality attributes, or advancing glycomics research, our glycoengineering and custom glycan modification services deliver the precision, scalability, and scientific rigor required to accelerate development and achieve superior biological outcomes.
References
