Structure of the carbohydrate moiety of human interferon-beta secreted by a recombinant Chinese hamster ovary cell line

Development of IFNβ-1a versions with reduced immunogenicity and full in vitro biological activity for the treatment of multiple sclerosis

IFNβ (recombinant interferon Beta) has been widely used for the treatment of Multiple sclerosis for the last four decades. Despite the human origin of the IFNβ sequence, IFNβ is immunogenic, and unwanted immune responses in IFNβ-treated patients may compromise its efficacy and safety in the clinic. In this study, we applied the DeFT (De-immunization of Functional Therapeutics) approach to producing functional, de-immunized versions of IFNβ-1a. Two de-immunized versions of IFNβ-1a were produced in CHO cells and designated as IFNβ-1a(VAR1) and IFNβ-1a(VAR2). First, the secondary and tertiary protein structures were analyzed by circular dichroism spectroscopy. Then, the variants were also tested for functionality. While IFNβ-1a(VAR2) showed similar in vitro antiviral activity to the original protein, IFNβ-1a(VAR1) exhibited 40% more biological potency. Finally, in vivo assays using HLA-DR transgenic mice revealed that the de-immunized variants showed a markedly reduced immunogenicity when compared to the originator.

Native Top-Down Mass Spectrometry Reveals a Role for Interfacial Glycans on Therapeutic Cytokine and Hormone Assemblies

Oligomerization and glycosylation modulate therapeutic glycoprotein stability and efficacy. The interplay between these two critical attributes on therapeutic glycoproteins, is however often hard to define. Here, we present a native top-down mass spectrometry (MS) approach to assess the glycosylation status of therapeutic cytokine and hormone assemblies and relate interfacial glycan occupancy to complex stability. We found that interfacial O-glycan stabilizes tumor necrosis factor-α trimer. On the contrary, interferon-β1a dimerization is independent of glycosylation. Moreover, we discovered a unique distribution of N-glycans on the follicle-stimulating hormone α subunit. We found that the interfacial N-glycan, at Asn52 of the α subunit, interacts extensively with the β subunit to regulate the dimer assembly. Overall, we have exemplified a method to link glycosylation with assembly status, for cytokines and hormones, critical for informing optimal stability and bioavailability.

Native Top-Down Mass Spectrometry Reveals a Role for Interfacial Glycans on Therapeutic Cytokine and Hormone Assemblies

Oligomerization and glycosylation modulate therapeutic glycoprotein stability and efficacy. The interplay between these two critical attributes on therapeutic glycoproteins, is however often hard to define. Here, we present a native top-down mass spectrometry (MS) approach to assess the glycosylation status of therapeutic cytokine and hormone assemblies and relate interfacial glycan occupancy to complex stability. We found that interfacial O-glycan stabilizes tumor necrosis factor-α trimer. On the contrary, interferon-β1a dimerization is independent of glycosylation. Moreover, we discovered a unique distribution of N-glycans on the follicle-stimulating hormone α subunit. We found that the interfacial N-glycan, at Asn52 of the α subunit, interacts extensively with the β subunit to regulate the dimer assembly. Overall, we have exemplified a method to link glycosylation with assembly status, for cytokines and hormones, critical for informing optimal stability and bioavailability.

Asn25 Deamidation as an Allosteric Tool to Increase IFNβ-1a Biological Activity

Interferon beta (IFNβ) is a well-known cytokine, belonging to the type I family, that exerts antiviral, immunomodulatory, and antiproliferative activity. It has been reported that the artificially deamidated form of recombinant IFNβ-1a at Asn25 position shows an increased biological activity. As a deepening of the previous study, the molecular mechanism underlying this biological effect was investigated in this work by combining experimental and computational techniques. Specifically, the binding to IFNAR1 and IFNAR2 receptors and the canonical pathway of artificially deamidated IFNβ-1a molecule were analyzed in comparison to the native form. As a result, a change in receptor affinity of deamidated IFNβ-1a with respect to the native form was observed, and to better explore this molecular interaction, molecular dynamics simulations were carried out. Results confirmed, as previously hypothesized, that the N25D mutation can locally change the interaction network of the mutated residue but also that this effect can be propagated throughout the molecule. In fact, many residues not involved in the interaction with IFNAR1 in the native form participate to the recognition in the deamidated molecule, enhancing the binding to IFNAR1 receptor and consequently an increase of signaling cascade activation. In particular, a higher STAT1 phosphorylation and interferon-stimulated gene expression was observed under deamidated IFNβ-1a cell treatment. In conclusion, this study increases the scientific knowledge of deamidated IFNβ-1a, deciphering its molecular mechanism, and opens new perspectives to novel therapeutic strategies.

Photo-induced radical thiol-ene chemistry: a versatile toolbox for peptide-based drug design

While the global market for peptide/protein-based therapeutics is witnessing significant growth, the development of peptide drugs remains challenging due to their low oral bioavailability, poor membrane permeability, and reduced metabolic stability. However, a toolbox of chemical approaches has been explored for peptide modification to overcome these obstacles. In recent years, there has been a revival of interest in photoinduced radical thiol-ene chemistry as a powerful tool for the construction of therapeutic peptides.

The Mechanistic Impact of N-Glycosylation on Stability, Pharmacokinetics, and Immunogenicity of Therapeutic Proteins

N-glycosylation is one of major post-translational modifications in nature, and it is essential for protein structure and function. As hydrophilic moieties of glycoproteins, N-glycans play important roles in protein stability. They protect the proteins against proteolytic degradation, aggregation, and thermal denaturation through maintaining optimal conformations. There are extensive evidences showing the involvement of N-glycans in the pharmacodynamics and pharmacokinetics of recombinant therapeutic proteins and antibodies. Highly sialylated complex-type glycans enable the longer serum half-lives of proteins against uptake through hepatic asialoglycoprotein receptor and mannose receptor for degradation in lysosomes. Moreover, the presence of nonhuman glycans results in clearance through pre-existing antibodies from serum and induces IgE-mediated anaphylaxis. N-glycans also facilitate or reduce the adverse immune responses of the proteins through interacting with multiple glycan-binding proteins, including those specific for mannose or mannose 6-phosphate. Due to the glycan impacts, a few therapeutic proteins were glycoengineered to improve the pharmacokinetics and stability. Thus, N-glycosylation should be extensively investigated and optimized for each individual protein for better efficacy and safety.

Site-specific derivatization of human interferon β-1a at lysine residues using microbial transglutaminase

Microbial transglutaminase (TGase) has been successfully used to produce site-specific protein conjugates derivatized at the level of glutamine (Gln) or lysine (Lys) residues with diverse applications. Here, we study the drug human interferon β-1a (IFN) as a substrate of TGase. The derivatization reaction was performed using carbobenzoxy-L-glutaminyl-glycine to modify Lys residues and dansylcadaverine for Gln residues. The 166 amino acids polypeptide chain of IFN β-1a contains 11 Lys and 11 Gln residues potential sites of TGase derivatization. By means of mass spectrometry analyses, we demonstrate the highly selective derivatization of this protein by TGase at the level of Lys115 and as secondary site at the level of Lys33, while no reactive Gln residue was detected. Limited proteolysis experiments were performed on IFN to determine flexible regions of the protein under physiological conditions. Interestingly, primary and secondary sites of limited proteolysis and of TGase derivatization occur at the same regions of the polypeptide chain, indicating that the extraordinary selectivity of the TGase-mediated reaction is dictated by the conformational features of the protein substrate. We envisage that the TGase-mediated derivatization of IFN can be used to produce interesting derivatives of this important therapeutic protein.

Effects of localized interactions and surface properties on stability of protein-based therapeutics

OBJECTIVES

Protein-based therapeutics garner significant attention because of exquisite specificity and limited side effects and are now being used to accomplish targeted delivery of small-molecule drugs. This review identifies and highlights individual chemical attributes and categorizes how site-specific changes affect protein stability based on published high-resolution molecular analyses.

KEY FINDINGS

Because it is challenging to determine the mechanisms by which the stability of large, complex molecules is altered and data are sparse, smaller, therapeutic proteins (insulin, erythropoietin, interferons) are examined alongside antibody data. Integrating this large pool of information with the limited available studies on antibodies reveals common mechanisms by which specific alterations affect protein structure and stability.

SUMMARY

Physical and chemical stability of therapeutic proteins and antibody drug conjugates (ADCs) is of critical importance because insufficient stability prevents molecules from making it to market. Individual moieties on/near the surface of proteins have substantial influence on structure and stability. Seemingly small, superficial modification may have far-reaching consequences on structure, conformational dynamics, and solubility of the protein, and hence physical stability of the molecule. Chemical modifications, whether spontaneous (e.g. oxidation, deamidation) or intentional, as with ADCs, may adversely impact stability by disrupting local surface properties or higher order protein structure.

The Role of Sialylated Glycans in Human Platelet Endothelial Cell Adhesion Molecule 1 (PECAM-1)-mediated Trans Homophilic Interactions and Endothelial Cell Barrier Function

Platelet Endothelial Cell Adhesion Molecule 1 (PECAM-1) is a major component of the endothelial cell intercellular junction. Previous studies have shown that PECAM-1 homophilic interactions, mediated by amino-terminal immunoglobulin homology domain 1, contribute to maintenance of the vascular permeability barrier and to its re-establishment following inflammatory or thrombotic insult. PECAM-1 glycans account for ∼30% of its molecular mass, and the newly solved crystal structure of human PECAM-1 immunoglobulin homology domain 1 reveals that a glycan emanating from the asparagine residue at position 25 (Asn-25) is located within the trans homophilic-binding interface, suggesting a role for an Asn-25-associated glycan in PECAM-1 homophilic interactions. In support of this possibility, unbiased molecular docking studies revealed that negatively charged α2,3 sialic acid moieties bind tightly to a groove within the PECAM-1 homophilic interface in an orientation that favors the formation of an electrostatic bridge with positively charged Lys-89, mutation of which has been shown previously to disrupt PECAM-1-mediated homophilic binding. To verify the contribution of the Asn-25 glycan to endothelial barrier function, we generated an N25Q mutant form of PECAM-1 that is not glycosylated at this position and examined its ability to contribute to vascular integrity in endothelial cell-like REN cells. Confocal microscopy showed that although N25Q PECAM-1 concentrates normally at cell-cell junctions, the ability of this mutant form of PECAM-1 to support re-establishment of a permeability barrier following disruption with thrombin was significantly compromised. Taken together, these data suggest that a sialic acid-containing glycan emanating from Asn-25 reinforces dynamic endothelial cell-cell interactions by stabilizing the PECAM-1 homophilic binding interface.

DC-SIGN and L-SIGN Are Attachment Factors That Promote Infection of Target Cells by Human Metapneumovirus in the Presence or Absence of Cellular Glycosaminoglycans

It is well established that glycosaminoglycans (GAGs) function as attachment factors for human metapneumovirus (HMPV), concentrating virions at the cell surface to promote interaction with other receptors for virus entry and infection. There is increasing evidence to suggest that multiple receptors may exhibit the capacity to promote infectious entry of HMPV into host cells; however, definitive identification of specific transmembrane receptors for HMPV attachment and entry is complicated by the widespread expression of cell surface GAGs. pgsA745 Chinese hamster ovary (CHO) cells are deficient in the expression of cell surface GAGs and resistant to HMPV infection. Here, we demonstrate that the expression of the Ca(2+)-dependent C-type lectin receptor (CLR) DC-SIGN (CD209L) or L-SIGN (CD209L) rendered pgsA745 cells permissive to HMPV infection. Unlike infection of parental CHO cells, HMPV infection of pgsA745 cells expressing DC-SIGN or L-SIGN was dynamin dependent and inhibited by mannan but not by pretreatment with bacterial heparinase. Parental CHO cells expressing DC-SIGN/L-SIGN also showed enhanced susceptibility to dynamin-dependent HMPV infection, confirming that CLRs can promote HMPV infection in the presence or absence of GAGs. Comparison of pgsA745 cells expressing wild-type and endocytosis-defective mutants of DC-SIGN/L-SIGN indicated that the endocytic function of CLRs was not essential but could contribute to HMPV infection of GAG-deficient cells. Together, these studies confirm a role for CLRs as attachment factors and entry receptors for HMPV infection. Moreover, they define an experimental system that can be exploited to identify transmembrane receptors and entry pathways where permissivity to HMPV infection can be rescued following the expression of a single cell surface receptor.

IMPORTANCE

On the surface of CHO cells, glycosaminoglycans (GAGs) function as the major attachment factor for human metapneumoviruses (HMPV), promoting dynamin-independent infection. Consistent with this, GAG-deficient pgaA745 CHO cells are resistant to HMPV. However, expression of DC-SIGN or L-SIGN rendered pgsA745 cells permissive to dynamin-dependent infection by HMPV, although the endocytic function of DC-SIGN/L-SIGN was not essential for, but could contribute to, enhanced infection. These studies provide direct evidence implicating DC-SIGN/L-SIGN as an alternate attachment factor for HMPV attachment, promoting dynamin-dependent infection via other unknown receptors in the absence of GAGs. Moreover, we describe a unique experimental system for the assessment of putative attachment and entry receptors for HMPV.

Formulation and stability of cytokine therapeutics

Cytokines are messenger proteins that regulate the proliferation and differentiation of cells and control immune responses. Interferons, interleukins, and growth factors have applications in cancer, autoimmune, and viral disease treatment. The cytokines are susceptible to chemical and physical instability. This article reviews the structure and stability issues of clinically used cytokines, as well as formulation strategies for improved stability. Some general aspects for identifying most probable stability concerns, selecting excipients, and developing stable cytokine formulations are presented. The vast group of cytokines offers possibilities for new biopharmaceuticals. The formulation approaches of the current cytokine products could facilitate development of new biopharmaceuticals.

Identification of oxidation sites and covalent cross-links in metal catalyzed oxidized interferon Beta-1a: potential implications for protein aggregation and immunogenicity

Oxidation via Cu(2+)/ascorbate of recombinant human interferon beta-1a (IFNβ1a) leads to highly immunogenic aggregates, however it is unknown which amino acids are modified and how covalent aggregates are formed. In the present work we mapped oxidized and cross-linked amino acid residues in aggregated IFNβ1a, formed via Cu(2+)/ascorbate catalyzed oxidation. Size exclusion chromatography (SEC) was used to confirm extensive aggregation of oxidized IFNβ1a. Circular dichroism and intrinsic fluorescence spectroscopy indicated substantial loss of secondary and tertiary structure, respectively. Derivatization with 4-(aminomethyl)benzenesulfonic acid was used to demonstrate, by fluorescence in combination with SEC, the presence of tyrosine (Tyr) oxidation products. High performance liquid chromatography coupled to electrospray ionization mass spectrometry of reduced, alkylated, and digested protein was employed to localize chemical degradation products. Oxidation products of methionine, histidine, phenylalanine (Phe), tryptophan, and Tyr residues were identified throughout the primary sequence. Covalent cross-links via 1,4- or 1,6-type addition between primary amines and DOCH (2-amino-3-(3,4-dioxocyclohexa-1,5-dien-1-yl)propanoic acid, an oxidation product of Phe and Tyr) were detected. There was no evidence of disulfide bridge, Schiff base, or dityrosine formation. The chemical cross-links identified in this work are most likely responsible for the formation of covalent aggregates of IFNβ1a induced by oxidation, which have previously been shown to be highly immunogenic.

Clinical testing for neutralizing antibodies to interferon-β in multiple sclerosis

Biopharmaceuticals are drugs which are based on naturally occurring proteins (antibodies, receptors, cytokines, enzymes, toxins), nucleic acids (DNA, RNA) or attenuated microorganisms. Immunogenicity of these agents has been commonly described and refers to a specific antidrug antibody response. Such immunogenicity represents a major factor impairing the efficacy of biopharmaceuticals due to biopharmaceutical neutralization. Indeed, clinical experience has shown that induction of antidrug antibodies is associated with a loss of response to biopharmaceuticals and also with hypersensitivity reactions. The first disease-specific agent licensed to treat multiple sclerosis (MS) was interferon-β (IFNβ). In its various preparations, it remains the most commonly used first-line agent. The occurrence of antidrug antibodies has been extensively researched in MS, particularly in relation to IFNβ. However, much controversy remains regarding the significance of these antibodies and incorporation of testing into clinical practice. Between 2% and 45% of people treated with IFNβ will develop neutralizing antibodies, and this is dependent on the specific drug and dosing regimen. The aim of this review is to discuss the use of IFNβ in MS, the biological and clinical relevance of anti-IFNβ antibodies (binding and neutralizing antibodies), the incorporation of testing in clinical practice and ongoing research in the field.

Chemical synthesis of homogeneous human glycosyl-interferon-β that exhibits potent antitumor activity in vivo

Chemical synthesis of homogeneous human glycoproteins exhibiting bioactivity in vivo has been a challenging task. In an effort to overcome this long-standing problem, we selected interferon-β and examined its synthesis. The 166 residue polypeptide chain of interferon-β was prepared by covalent condensation of two synthetic peptide segments and a glycosylated synthetic peptide bearing a complex-type glycan of biological origin. The peptides were covalently condensed by native chemical ligation. Selective desulfurization followed by deprotection of the two Cys(Acm) residues gave the target full-length polypeptide chain of interferon-β bearing either a complex-type sialyl biantennary oligosaccharide or its asialo form. Subsequent folding with concomitant formation of the native disulfide bond afforded correctly folded homogeneous glycosyl-interferon-β. The chemically synthesized sialyl interferon-β exhibited potent antitumor activity in vivo.

Development of resistance to biologic therapies with reference to IFN-β

All biotherapeutics have the potential to generate anti-drug antibodies (ADAs) in patients. The main factors leading to an immune response are thought to be product, treatment and patient related. In this review, reasons for the formation of ADAs, and particularly neutralizing antibodies (NAbs), are considered, with a focus on IFN-β as a well-studied example. The time course for the production of NAbs, the measurement of NAbs, the defining of IFN-β responders and non-responders, the implications for disease progression in patients, and future methods for avoiding the production of ADAs and of tolerizing patients are considered.

N-linked glycosylation facilitates sialic acid-independent attachment and entry of influenza A viruses into cells expressing DC-SIGN or L-SIGN

It is widely recognized that sialic acid (SA) can mediate attachment of influenza virus to the cell surface, and yet the specific receptors that mediate virus entry are not known. For many viruses, a definitive demonstration of receptor function has been achieved when nonpermissive cells are rendered susceptible to infection following transfection of the gene encoding a putative receptor. For influenza virus, such approaches have been confounded by the abundance of SA on mammalian cells so that it has been difficult to identify cell lines that are not susceptible to infection. We examined influenza virus infection of Lec2 Chinese hamster ovary (CHO) cells, a mutant cell line deficient in SA. Lec2 CHO cells were resistant to influenza virus infection, and stable cell lines expressing either DC-SIGN or L-SIGN were generated to assess the potential of each molecule to function as SA-independent receptors for influenza A viruses. Virus strain BJx109 (H3N2) bound to Lec2 CHO cells expressing DC-SIGN or L-SIGN in a Ca(2+)-dependent manner, and transfected cells were susceptible to virus infection. Treatment of Lec2-DC-SIGN and Lec2-L-SIGN cells with mannan, but not bacterial neuraminidase, blocked infection, a finding consistent with SA-independent virus attachment and entry. Moreover, virus strain PR8 (H1N1) bears low levels of mannose-rich glycans and was inefficient at infecting Lec2 CHO cells expressing either DC-SIGN or L-SIGN, whereas other glycosylated H1N1 subtype viruses could infect cells efficiently. Together, these data indicate that human C-type lectins (DC-SIGN and L-SIGN) can mediate attachment and entry of influenza viruses independently of cell surface SA.

Aggregated recombinant human interferon Beta induces antibodies but no memory in immune-tolerant transgenic mice

Purpose

To study the influence of protein aggregation on the immunogenicity of recombinant human interferon beta (rhIFNβ) in wild-type mice and transgenic, immune-tolerant mice, and to evaluate the induction of immunological memory.

Methods

RhIFNβ-1b and three rhIFNβ-1a preparations with different aggregate levels were injected intraperitoneally in mice 15× during 3 weeks, and the mice were rechallenged with rhIFNβ-1a. The formation of binding (BABs) and neutralizing antibodies (NABs) was monitored.

Results

Bulk rhIFNβ-1a contained large, mainly non-covalent aggregates and stressed rhIFNβ-1a mainly covalent, homogeneous (ca. 100 nm) aggregates. Reformulated rhIFNβ-1a was essentially aggregate-free. All products induced BABs and NABs in wild-type mice. Immunogenicity in the transgenic mice was product dependent. RhIFNβ-1b showed the highest and reformulated rhIFNβ-1a the lowest immunogenicity. In contrast with wild-type mice, transgenic mice did not show NABs, nor did they respond to the rechallenge.

Conclusions

The immunogenicity of the products in transgenic mice, unlike in wild-type mice, varied. In the transgenic mice, neither NABs nor immunological memory developed. The immunogenicity of rhIFNβ in a model reflecting the human immune system depends on the presence and the characteristics of aggregates.

Identification of genes encoding N-glycan processing beta-N-acetylglucosaminidases in Trichoplusia ni and Bombyx mori: Implications for glycoengineering of baculovirus expression systems

Glycoproteins produced by non-engineered insects or insect cell lines characteristically bear truncated, paucimannose N-glycans in place of the complex N-glycans produced by mammalian cells. A key reason for this difference is the presence of a highly specific N-glycan processing beta-N-acetylglucosaminidase in insect, but not in mammalian systems. Thus, reducing or abolishing this enzyme could enhance the ability of glycoengineered insects or insect cell lines to produce complex N-glycans. Of the three insect species routinely used for recombinant glycoprotein production, the processing beta-N-acetylglucosaminidase gene has been isolated only from Spodoptera frugiperda. Thus, the purpose of this study was to isolate and characterize the genes encoding this important processing enzyme from the other two species, Bombyx mori and Trichoplusia ni. Bioinformatic analyses of putative processing beta-N-acetylglucosaminidase genes isolated from these two species indicated that each encoded a product that was, indeed, more similar to processing beta-N-acetylglucosaminidases than degradative or chitinolytic beta-N-acetylglucosaminidases. In addition, over-expression of each of these genes induced an enzyme activity with the substrate specificity characteristic of processing, but not degradative or chitinolytic enzymes. Together, these results demonstrated that the processing beta-N-acetylglucosaminidase genes had been successfully isolated from Trichoplusia ni and Bombyx mori. The identification of these genes has the potential to facilitate further glycoengineering of baculovirus-insect cell expression systems for the production of glycosylated proteins.

Glycosylation of therapeutic proteins: an effective strategy to optimize efficacy

During their development and administration, protein-based drugs routinely display suboptimal therapeutic efficacies due to their poor physicochemical and pharmacological properties. These innate liabilities have driven the development of molecular strategies to improve the therapeutic behavior of protein drugs. Among the currently developed approaches, glycoengineering is one of the most promising, because it has been shown to simultaneously afford improvements in most of the parameters necessary for optimization of in vivo efficacy while allowing for targeting to the desired site of action. These include increased in vitro and in vivo molecular stability (due to reduced oxidation, cross-linking, pH-, chemical-, heating-, and freezing-induced unfolding/denaturation, precipitation, kinetic inactivation, and aggregation), as well as modulated pharmacodynamic responses (due to altered potencies from diminished in vitro enzymatic activities and altered receptor binding affinities) and improved pharmacokinetic profiles (due to altered absorption and distribution behaviors, longer circulation lifetimes, and decreased clearance rates). This article provides an account of the effects that glycosylation has on the therapeutic efficacy of protein drugs and describes the current understanding of the mechanisms by which glycosylation leads to such effects.

Carbohydrate analysis throughout the development of a protein therapeutic

This review discusses the challenges involved in the characterization of the glycosylation of therapeutic glycoproteins. The focus is on methods that are most commonly used in regulatory filings and lot release testing of therapeutic glycoproteins. The different types of assays for carbohydrate analysis are reviewed, including the distinction between assays appropriate for lot release or better suited to testing during early drug development or in-depth characterization of the glycosylation. Characteristics of the glycoprotein and production process that should be considered when determining the amount of testing, the number of different methods to employ and when the testing should be performed during development of protein therapeutics is also discussed.

Effects of glycosylation on the stability of protein pharmaceuticals

In recent decades, protein-based therapeutics have substantially expanded the field of molecular pharmacology due to their outstanding potential for the treatment of disease. Unfortunately, protein pharmaceuticals display a series of intrinsic physical and chemical instability problems during their production, purification, storage, and delivery that can adversely impact their final therapeutic efficacies. This has prompted an intense search for generalized strategies to engineer the long-term stability of proteins during their pharmaceutical employment. Due to the well known effect that glycans have in increasing the overall stability of glycoproteins, rational manipulation of the glycosylation parameters through glycoengineering could become a promising approach to improve both the in vitro and in vivo stability of protein pharmaceuticals. The intent of this review is therefore to further the field of protein glycoengineering by increasing the general understanding of the mechanisms by which glycosylation improves the molecular stability of protein pharmaceuticals. This is achieved by presenting a survey of the different instabilities displayed by protein pharmaceuticals, by addressing which of these instabilities can be improved by glycosylation, and by discussing the possible mechanisms by which glycans induce these stabilization effects.

The Golgi CMP-sialic acid transporter: A new CHO mutant provides functional insights

A CHO mutant line, MAR-11, was isolated using a cytotoxic lectin, Maackia amurensis agglutinin (MAA). This mutant has decreased levels of cell surface sialic acid relative to both wild-type CHO-K1 and Lec2 mutant CHO cells. The CMP-sialic acid transporter (CMP-SAT) gene in the MAR-11 mutant cell has a C-T mutation that results in a premature stop codon. As a result, MAR-11 cells express a truncated version of CMP-SAT which contains only 100 amino acids rather than the normal CMP-SAT which contains 336 amino acids. Biochemical analyses indicate that recombinant interferon-gamma (IFN-gamma) produced by the mutant cells lack sialic acid. Using MAR-11 as host cells, an EPO/IEF assay for the structure-function study of CMP-SAT was developed. This assay seems more sensitive than previous assays that were used to analyze sialylation in Lec2 cells. Cotransfection of constructs that express CMP-SAT into MAR-11 cells completely converted the recombinant EPO to a sialylation pattern that is similar to the EPO produced by the wild-type CHO cells. Using this assay, we showed that CMP-SAT lacking C-terminal 18 amino acids from the cytosolic tail was able to allow high levels of EPO sialylation. Substitution of the Gly residues with Ile in three different transmembrane domains of CMP-SAT resulted in dramatic decreases in transporter's activity. The CMP-SAT only lost partial activity if the same Gly residues were substituted with Ala, suggesting that the lack of side chain in Gly residues in the transmembrane domains is essential for transport activity.

The function of the human interferon-beta 1a glycan determined in vivo

Recombinant human interferon-beta (rhIFN-beta) is the leading therapeutic intervention shown to change the cause of relapsing-remitting multiple sclerosis, and both a nonglycosylated and a significantly more active glycosylated variant of rhIFN-beta are used in treatment. This study investigates the function of the rhIFN-beta1a glycan moiety and its individual carbohydrate residues, using the myxovirus resistance (Mx) mRNA as a biomarker in Mx-congenic mice. We showed that the Mx mRNA level in blood leukocytes peaked 3 h after s.c. administration of rhIFN-beta1a. In addition, a clear dose-response relationship was confirmed, and the Mx response was shown to be receptor-mediated. Using specific glycosidases, different glycosylation analogs of rhIFN-beta1a were obtained, and their activities were determined. The glycosylated rhIFN-beta1a showed significantly higher activity than its deglycosylated counterpart, due to a protein stabilization/solubilization effect of the glycan. It is interesting to note that the terminating sialic acids were essential for these effects. Conclusively, the structure/bioactivity relationship of rhIFN-beta1a was determined in vivo, and it provided a novel insight into the role of the rhIFN-beta1a glycan and its carbohydrate residues. The possibilities of improving the pharmacological properties of rhIFN-beta1a using glycoengineering are discussed.

Neutralising antibodies to interferon β in multiple sclerosis

Interferon beta (IFNbeta) therapy for multiple sclerosis (MS) is associated with a potential for the development of neutralising antibodies (NAbs) that negatively affect therapy. Several factors influence the development of NAbs, such as lack of complete sequence homology with the endogenous IFNbeta sequence, frequency of administration, level of dose and formulation of IFNbeta. Taken together, the evidence that NAb status reduces clinical efficacy in MS patients is strong. Standardised assays for NAbs are lacking, and titres vary over time. NAb testing is a critical component of care for MS patients because it provides information on one of the most important factors determining clinical responsiveness to IFNbeta therapy. This expert panel report attempts to move the field towards resolution of the remaining issues and considers several aspects of NAbs, including their clinical relevance, factors influencing immunogenicity, assays to quantify NAbs and the definition of clinically relevant titres.

Post-translational modifications in the context of therapeutic proteins

The majority of protein-based biopharmaceuticals approved or in clinical trials bear some form of post-translational modification (PTM), which can profoundly affect protein properties relevant to their therapeutic application. Whereas glycosylation represents the most common modification, additional PTMs, including carboxylation, hydroxylation, sulfation and amidation, are characteristic of some products. The relationship between structure and function is understood for many PTMs but remains incomplete for others, particularly in the case of complex PTMs, such as glycosylation. A better understanding of such structural-functional relationships will facilitate the development of second-generation products displaying a PTM profile engineered to optimize therapeutic usefulness.

Production and glycosylation of recombinant beta-interferon in suspension and cytopore microcarrier cultures of CHO cells

Microcarriers are suitable for high-density cultures of cells requiring surface attachment and also offer the advantage of easy media removal for product recovery. We have used the macroporous microcarriers Cytopore 1 and 2 for the growth of CHO cells producing recombinant human beta-interferon (beta-IFN) in stirred batch cultures. Although these cells may grow in suspension, in the presence of Cytopore microcarriers they become entrapped in the inner bead matrix where they can be maintained at high densities. Cell growth rates were reduced in microcarrier cultures compared to suspension cultures. However, the beta-IFN yield was up to 3-fold greater as a result of an almost 5-fold higher specific productivity. Maximum productivity was found in cultures containing 1.0 mg/mL of Cytopore 1 or 0.5 mg/mL of Cytopore 2 with a cell/bead ratio of 1029 and 822, respectively. Beta-IFN molecules aggregated in the later stages of all cultures, causing a decrease in response by ELISA. However, the degree of aggregation was significantly less in the microcarrier cultures. The N-linked glycans from beta-IFN were isolated and analyzed by normal phase HPLC. There was no apparent difference in the profile of glycans obtained from each of the suspension and Cytopore culture systems. This suggests that Cytopore microcarriers may be useful in bioprocess development for enhanced recombinant glycoprotein production without affecting the glycosylation profile of the protein.

Metabolic Functionalization of Recombinant Glycoproteins

Glycoproteins are essential for cellular communication and are the most rapidly growing class of therapeutic agents. Chemical modification of glycoproteins has been employed to improve their in vivo efficacy or to label them for detection. Methods for the controlled derivatization of glycoproteins are presently limited by the repertoire of natural amino acid side chain and carbohydrate functionalities. Here, we use metabolic oligosaccharide engineering to introduce a bioorthogonal functional group, the azide, into cellular and recombinant glycoproteins for subsequent chemical elaboration via Staudinger ligation. As most therapeutic glycoproteins are sialylated and require this saccharide for optimal pharmacokinetics, we targeted sialic acid as a host for azides using N-azidoacetylmannosamine (ManNAz) as a biosynthetic precursor. Metabolic conversion of ManNAz to N-azidoacetylsialic acid (SiaNAz) within membrane-bound and secreted glycoproteins was quantified in a variety of cell types. SiaNAz was found to comprise between 4% and 41% of total sialosides, depending on the system. Metabolic labeling of recombinant interferon-beta and GlyCAM-Ig was achieved, demonstrating the utility of the method for functionalizing N-linked and O-linked glycoproteins of therapeutic interest. More generally, the generation of recombinant glycoproteins containing chemical handles within their glycans provides a means for studying their behavior and for improving their in vivo efficacy.

Expression, purification, and characterization of rat interferon-β, and preparation of an N-terminally PEGylated form with improved pharmacokinetic parameters

To identify potential new clinical uses and routes of administration for human interferon-beta-1a (IFN-beta-1a), we have developed an expression and purification procedure for the preparation of highly purified rat interferon-beta (IFN-beta) suitable for testing in rat models of human disease. An expression vector containing the rat IFN-beta signal sequence and structural gene was constructed and transfected into Chinese hamster ovary (CHO) cells. The protein was purified from CHO cell conditioned medium and purified to > 99.5% purity using standard chromatographic techniques. Analytical characterization indicated that the protein was a heavily glycosylated monomeric protein, with two of the four predicted N-glycosylation sites occupied. Analysis of the attached oligosaccharides showed them to be a complex mixture of bi-antennary, tri-antennary, and tetra-antennary structures with a predominance of sialylated tri-antennary and tetra-antennary structures. Peptide mapping, N-terminal sequencing, and mass spectrometry confirmed the identity and integrity of the purified protein. The purified protein had a specific activity of 2.1x10(8)U/mg when assayed on rat RATEC cells, which is similar in magnitude to the potencies observed for murine IFN-beta and human IFN-beta-1a assayed on murine and human cells, respectively. We also prepared an N-terminally PEGylated form of rat IFN-beta in which a 20 kDa methoxy polyethylene glycol (PEG)-propionaldehyde was attached to the N-terminal alpha-amino group of Ile-1. The PEGylated protein, which retained essentially full in vitro antiviral activity, had improved pharmacokinetic parameters in rats as compared to the unmodified protein. Both the unmodified and PEGylated forms of rat IFN-beta will be useful for testing in rat models of human disease.

N-linked glycan structures of mouse interferon-β produced by Bombyx mori larvae

The full-length mouse interferon-beta (mIFN-beta) cDNA, including the secretion signal peptide coding region under control of the polyhedrin promoter, was introduced into Bombyx mori nucleopolyhedrovirus (BmNPV). Recombinant mIFN-beta (rmIFN-beta) was accumulated in the haemolymph of infected silkworm larvae. Western blot analysis showed isoforms of rmIFN-beta, suggesting that rmIFN-beta is glycosylated. The glycan structures of purified rmIFN-beta were determined. The N-glycans were liberated by hydrazinolysis and the resulting oligosaccharides were labeled with 2-aminopyridine. The pyridylaminated (PA) glycans were purified by gel filtration, reversed-phase HPLC, and size-fractionation HPLC. The structures of the PA-sugar chains were identified by a combination of two-dimensional PA-sugar chain mapping, MS analysis, and exoglycosidase digestions.

Monitoring biodistribution of glycoproteins with modified sugar chains

Natural human interferon (hIFN)-gamma has mainly biantennary complex-type sugar chains. Previously, we successfully remodeled its sugar chain structure into: (a) highly branched types; or (b) highly sialylated types, by overexpression of: (a) N-acetylglucosaminyltransferase (GnT)-IV and/or GnT-V; or (b) sialyltransferases, in Chinese hamster ovary (CHO) cells. In addition, we prepared asialo hIFN-gammas by treatment with sialidase in vitro. In the present study, we assessed the bioactivity of remodeled hIFN-gamma in terms of antiviral activity, anticellular activity, and biodistribution. Structural changes to the sugar chains did not have a significant influence on the antiviral and anticellular activities of hIFN-gamma, although the attachment of the sugar chain itself affected both activities. However, the biodistribution differed significantly; the number of exposed galactose residues was the major determinant of the specific distribution to the liver and blood clearance rate of hIFN-gamma. This phenomenon was considered to be mediated by the hepatic asialoglycoprotein receptor (ASGP-R), and we showed a linear, not exponential, enhancement of the distribution to the liver with an increase in the number of exposed galactose residues. We also confirmed this tendency using fibroblast growth factor (FGF). Our observation is not the same as the "glycoside cluster effect." We thus provide important information on the character of modified recombinant glycoproteins.

Application of the StrOligo algorithm for the automated structure assignment of complex N-linked glycans from glycoproteins using tandem mass spectrometry

Oligosaccharides associated with proteins are known to give these molecules specific conformations and functions. Analysis of proteins would not be complete without studying the glycans. However, high-throughput techniques in proteomics will soon overwhelm the current capacity of methods if no automation is incorporated into glycomics. New capabilities of the StrOligo algorithm introduced for this purpose (Ethier et al., Rapid Commun. Mass Spectrom., 2002; 16: 1743) will be discussed here. Experimental tandem mass spectra were acquired to test the algorithm using a hybrid quadrupole-time-of-flight (QqTOF) instrument with a matrix-assisted laser desorption/ionization (MALDI) source. The samples were N-linked oligosaccharides from monoclonal antibody IgG, beta interferon and fetuin, detached by enzymatic deglycosylation and labeled at the reducing end. Improvements to the program were made in order to reduce the need for user intervention. StrOligo strips the spectra down to monoisotopic peaks only. The algorithm first builds a relationship tree, accounting for each observed loss of a monosaccharide moiety, and then analyzes the tree and proposes possible structures from combinations of adducts and fragment ion types. A score, which reflects agreement with experimental results, is then given to each proposed structure. The program then decides which combination is the best one and labels relevant peaks in the experimental mass spectrum using a modified nomenclature. The usefulness of the algorithm has been demonstrated by assigning structures to several glycans released from glycoproteins. The analysis was completed in less than 2 minutes for any glycan, which is a substantial improvement over manual interpretation.

Structural characterization of oligosaccharides in recombinant soluble human interferon receptor 2 using fluorophore-assisted carbohydrate electrophoresis

The N-linked oligosaccharide profiles (banding patterns in gels) and structures of recombinant soluble human interferon receptor 2 (r-shIFNAR2) were determined using fluorophore-assisted carbohydrate electrophoresis (FACE, Glyko, Novato, CA). The method involves releasing N-linked oligosaccharide moieties from a glycoprotein by digestion with peptide-N glycanase (PNGase F), labeling the released oligosaccharides with the fluorescent dye 8-aminonaphthalene-1,3,6-trisulfonate (ANTS), and separating the labeled oligosaccharides by gel electrophoresis. The isolated oligosaccharides in the bands from the profiling gels can then be sequenced using exoglycosidases to reveal the oligosaccharide structures. The oligosaccharide profile of r-shIFNAR2 consists of at least nine oligosaccharide bands. The relative amount of oligosaccharide in each band can vary, depending on the culture conditions of the source cells. FACE structural analysis shows that r-shIFNAR2 contains only core-fucosylated N-linked oligosaccharides, most of which are fully sialylated (approximately 92%). The major types and relative amounts of the oligosaccharides from a representative sample are: disialylated, galactosylated, biantennary (15%); trisialylated, galactosylated, triantennary (19%), tetrasialylated, galactosylated, tetraantennary (30%), and N-acetyllactosamine-containing higher-order oligosaccharides including tri-, tetra-, and pentaantennary (28%). The remaining oligosaccharides are not fully sialylated and/or not fully galactosylated di-, tri-, and tetraantennary structures (approximately 5%) and unidentified structures (approximately 3%). A method for determining the types and structures of the N-acetyllactosamine containing oligosaccharides is also reported in this study.

Structural analysis of modified forms of recombinant IFN-beta produced under stress-simulating conditions

The present study focused on the investigation of the chemical stability of recombinant human interferon-beta (rhIFN-beta) tested in vitro by chemical treatments that simulate stress-induced conditions that may occur during handling, storage or ageing of protein samples. Mild oxidation and/or alkylation of the recombinant protein showed that the four methionines occurring in the interferon displayed different chemical susceptibility in that Met36 and Met117 were fully modified, whereas Met1 showed only little modification and Met62 was completely resistant. Moreover, incubation of rhIFN-beta under alkaline conditions resulted in the formation of a covalent dimeric species stabilised by an intermolecular disulphide bridge involving the free SH group of Cys17 from each polypeptide chain. Analysis of biological activity of the different IFN-beta derivatives showed that rhIFN-beta fully retains its specific activity following mild oxidation treatments whereas reaction with a high concentration of alkylating agents or incubation under alkaline conditions strongly reduce its specific antiviral activity.

Study of the sugar chains of recombinant human amyloid precursor protein produced by Chinese hamster ovary cells

The N- and O-glycans of recombinant amyloid precursor protein (APP), purified from Chinese hamster ovary cells transfected with the human 695-amino acid form of APP, were separately released by hydrazinolysis under different conditions. The reducing ends of the released N- and O-glycans were reduced with NaB3H4 and derivatized with 2-aminobenzamide (2AB), respectively. After acidic N-glycans were obtained by anion-exchange column chromatography, these were converted to neutral oligosaccharides by sialidase digestion, demonstrating that their acidic nature was entirely due to sialylation. The sialidase-treated N-glycans were then fractionated by lectin column chromatography and their structures were determined by linkage-specific sequential exoglycosidase digestion. These results demonstrated that recombinant APP has bi- and triantennary complex type N-glycans with fucosylated and nonfucosylated trimannosyl cores. In a similar fashion, the 2AB-labeled O-glycans derived from APP were determined to be mono- and disialylated core type 1 structures. Taken together, these results indicate that recombinant APP has sialylated bi- and triantennary N-glycans with fucosylated and nonfucosylated cores and sialylated O-glycans with core type 1 structures.

Enhanced inhibition of hepatitis B virus production by asialoglycoprotein receptor-directed interferon

Most chronic carriers of hepatitis B virus (HBV) do not respond to interferon (IFN) treatment. This limitation of IFN therapy may be due in part to scant expression of IFN receptor in the liver. Because the asialoglycoprotein (ASGP) receptor is specifically expressed in the liver at high density, the ASGP receptor-binding domain was generated within an N-glycosylated human IFN-beta molecule by the removal of sialic acid to direct this cytokine to the liver. This modified IFN (asialo-IFN-beta) demonstrated greater inhibition of HBV production in ASGP receptor-positive human liver cells transfected with a replication-competent HBV construct than did conventional IFN-alpha or IFN-beta. Furthermore, the enhanced antiviral effect of asialo-IFN-beta was supported by induction of the 2'-5' oligoadenylate synthetase, an indicator of IFN activity, at a level significantly higher than that produced by conventional IFN-beta. Moreover, mouse asialo-IFN-beta profoundly reduced viremia in vivo in HBV-transfected athymic nude mice, in contrast to conventional IFN-beta, which had no substantial effect. These experiments demonstrate that directing IFN to ASGP receptor facilitates its signaling in the liver and augments its antiviral effect, and is therefore useful in overcoming the limited antiviral effect of conventional IFNs.

Stable expression of human alpha-2,6-sialyltransferase in Chinese hamster ovary cells: functional consequences for human erythropoietin expression and bioactivity

Hamster cell lines are common hosts for recombinant protein production, e.g. erythropoietin (Epo). Terminal sialylation of native human proteins is characteristically in both alpha-2,3 and alpha-2,6 linkage to galactose at the termini of N-linked oligosaccharides but only in alpha-2,3 linkage in recombinant proteins expressed in hamster cells which do not express alpha-2, 6-sialyltransferase (ST6GalI) (EC 2.4.99.1). This difference could alter the bioactivity of certain recombinant proteins. Chinese hamster ovary (CHO) cells stably transfected with human ST6GalI cDNA linked to the hamster metallothionein II promoter expressed highly inducible authentic ST6GalI activity. Untransfected CHO cells and CHO cells stably expressing ST6GalI cDNA when transfected with a human Epo cDNA expression cassette secreted immunoreactive Epo. Human Epo from singly transfected Pro-5 CHO cells induced significant reticulocytosis (7.00+/-1.58%; mean+/-S.D. % reticulocytes; control conditioned medium 3.04+/-1.29%; P<0.0024), whereas Epo from Pro-5 cells coexpressing ST6GalI elicited a more modest reticulocytosis (4.62+/-1.02%). Thus for recombinant human Epo, engineering CHO cells to express ST6GalI activity does not enhance Epo bioactivity in vivo in mice. The availability of CHO cells that express high levels of ST6GalI activity now enables systematic studies to determine the functional requirement for this form of sialylation in recombinant human proteins.

The crystal structure of human interferon beta at 2.2-A resolution

Type I interferons (IFNs) are helical cytokines that have diverse biological activities despite the fact that they appear to interact with the same receptor system. To achieve a better understanding of the structural basis for the different activities of alpha and beta IFNs, we have determined the crystal structure of glycosylated human IFN-beta at 2.2-A resolution by molecular replacement. The molecule adopts a fold similar to that of the previously determined structures of murine IFN-beta and human IFN-alpha2b but displays several distinct structural features. Like human IFN-alpha2b, human IFN-beta contains a zinc-binding site at the interface of the two molecules in the asymmetric unit, raising the question of functional relevance for IFN-beta dimers. However, unlike the human IFN-alpha2b dimer, in which homologous surfaces form the interface, human IFN-beta dimerizes with contact surfaces from opposite sides of the molecule. The relevance of the structure to the effects of point mutations in IFN-beta at specific exposed residues is discussed. A potential role of ligand-ligand interactions in the conformational assembly of IFN receptor components is discussed.

Elongation of N-acetyllactosamine repeats in diantennary oligosaccharides

Glycosylated [Asn22]lysozyme has been shown to contain N-acetyllactosamine repeats when expressed in chinese hamster ovary (CHO) cells. We find that the major portion of N-acetyllactosamine repeats are associated with diantennary oligosaccharides. In Lec2 CHO cells, which are deficient in sialylation, glycosylated lysozyme is synthesized with increased contents of N-acetyllactosamine repeats terminating in beta-galactosyl residues. In the Lec2 cells and the parental CHO cell line, Pro 5, only a minor portion of the oligosaccharides in lysozyme are of the triantennary type. Previously, it has been shown that the synthesis of N-acetyllactosamine repeats in Asn-linked oligosaccharides is enhanced by an increase in the activity of the elongating beta-N-acetylglucosaminyl transferase and by the synthesis of beta-1,6-linked antennae. The results with glycosylated lysozyme suggest that glycoproteins bearing diantennary oligosaccharides can contain several N-acetyllactosamine repeats and that the number of the latter can be increased by decreasing the activity of the capping sialyl transferases.

Characterization of changes in the glycosylation pattern of recombinant proteins from BHK-21 cells due to different culture conditions

The N-glycosylation patterns of a genetically engineered human interleukin-2 variant glycoprotein (IL-Mu6), produced by BHK-21 cells from long-term suspension and microcarrier cultures in the presence and absence of fetal calf serum were compared. IL-Mu6 was used as a model protein in studying the effect of different controlled cell culture conditions on the expression of N-glycans in recombinant glycoproteins. IL-Mu6 contains a single amino acid substitution (Glu100<==>Asn) generating a potential N-glycosylation recognition site (Asn100-Xxx-Thr/Ser) in addition to the natural O-glycosylation at position Thr3. Parallel cell cultivations were carried out in two continuously perfused 2.5-liter stirred bioreactors under defined culture conditions. Major differences were found in the glycoprotein products obtained during these different cultivation conditions. Serum-free cultures resulted in a higher level of terminal sialylation and proximal alpha 1-6 fucosylation. The ratio of O- to N-glycans as well as the amount of nonglycosylated product and the antennarity of N-linked carbohydrates in the model protein exhibited major differences depending on the presence or absence of serum, the condition of growth and the cultivation procedure.