Glycan engineering reveals interrelated effects of terminal galactose and core fucose on antibody-dependent cell-mediated cytotoxicity.

Antibody-dependent cell-mediated cytotoxicity (ADCC) has been identified as one of the potentially critical effector functions underlying the clinical efficacy of some therapeutic immunoglobin G1 (IgG1) antibodies. It has been well established that higher levels of afucosylated N-linked glycan structures on the Fc region enhance the IgG binding affinity to the FcγIIIa receptor and lead to increased ADCC activity. However, whether terminal galactosylation of an IgG1 impacts its ADCC activity is less understood. Here, we used a new strategy for glycan enrichment and remodeling to study the impact of terminal galactose on ADCC activity for therapeutic IgG1s. Our results indicate that the degree of influence of terminal galactose on in vitro ADCC activity depends on the presence or absence of the core fucose, which is typically linked to the first N-acetyl glucosamine residue of an N-linked glycosylation core structure. Specifically, terminal galactose on afucosylated IgG1 mAbs enhanced ADCC activity with impact coefficients (ADCC%/Gal%) more than 20, but had minimal influence on ADCC activity on fucosylated structures with impact coefficient in the range of 0.1-0.2. Knowledge gained here can be used to guide product and process development activities for biotherapeutic antibodies that require effector function for efficacy, and also highlight the complexity in modulating the immune response through N-linked glycosylation of antibodies.

Use of In Vitro Assays to Assess Immunogenicity Risk of Antibody-Based Biotherapeutics.

An In Vitro Comparative Immunogenicity Assessment (IVCIA) assay was evaluated as a tool for predicting the potential relative immunogenicity of biotherapeutic attributes. Peripheral blood mononuclear cells from up to 50 healthy naïve human donors were monitored up to 8 days for T-cell proliferation, the number of IL-2 or IFN-γ secreting cells, and the concentration of a panel of secreted cytokines. The response in the assay to 10 monoclonal antibodies was found to be in agreement with the clinical immunogenicity, suggesting that the assay might be applied to immunogenicity risk assessment of antibody biotherapeutic attributes. However, the response in the assay is a measure of T-cell functional activity and the alignment with clinical immunogenicity depends on several other factors. The assay was sensitive to sequence variants and could differentiate single point mutations of the same biotherapeutic. Nine mAbs that were highly aggregated by stirring induced a higher response in the assay than the original mAbs before stirring stress, in a manner that did not match the relative T-cell response of the original mAbs. In contrast, mAbs that were glycated by different sugars (galactose, glucose, and mannose) showed little to no increase in response in the assay above the response to the original mAbs before glycation treatment. The assay was also used successfully to assess similarity between multiple lots of the same mAb, both from the same manufacturer and from different manufacturers (biosimilars). A strategy for using the IVCIA assay for immunogenicity risk assessment during the entire lifespan development of biopharmaceuticals is proposed.

Evaluating Immunogenicity Risk Due to Host Cell Protein Impurities in Antibody-Based Biotherapeutics.

A potential risk factor for immunogenicity of a biotherapeutic is the low levels of host cell protein (HCP) impurities that remain in the product following the purification process. During process development, significant attention has been devoted to removing HCPs due to their potential safety risk. Samples from different purification steps of several monoclonal antibodies (mAbs) purified by one type of platform were evaluated for their residual Chinese Hamster Ovary (CHO) cell-derived HCP content. HCPs in both in-process (high levels of HCP) and highly purified (low levels of HCP) samples were identified and quantitated by proteomic analysis via mass spectrometry. The responses to HCPs were evaluated in an in vitro assay using PBMC from a population of healthy and disease state individuals. Results indicated that samples with up to 4000 ppm HCP content (levels 200 times greater than the drug substance) did not pose a higher immunogenicity risk than highly purified mAb samples. As an orthogonal method to predict immunogenicity risk, in silico algorithms that probe amino acid sequence for foreign epitope content were used to evaluate over 20 common HCPs (identified in the different mAb samples). Only a few HCPs were identified as high risk by the algorithms; however, the in vitro assay results indicated that the concentration of these HCPs from in-process biotherapeutic mAb samples was not sufficient to stimulate an immune response. This suggests that high levels of HCP in mAb biotherapeutics purified by this type of platform do not increase the potential risk of immunogenicity of these molecules. Insights from these studies can be applied to HCP control and risk assessment strategies.

Characterization of the co-elution of host cell proteins with monoclonal antibodies during protein A purification.

Protein A chromatography is commonly used as the initial step for purifying monoclonal antibody biotherapeutics expressed in mammalian tissue culture cells. The purpose of this step, as well as later chromatography steps, is, in part, to remove host cell proteins (HCPs) and other related impurities. Understanding the retention mechanism for the subset of HCPs retained during this step is of great interest to monoclonal antibody (mAb) process developers because it allows formation of a guided HCP clearance strategy. However, only limited information is available about the specific HCPs that co-purify with mAbs at this step. In this study, a comprehensive comparison of HCP subpopulations that associated with 15 different mAbs during protein A chromatography was conducted by a 2D-LC-HDMS(E) approach. We found that a majority of CHO HCPs binding to and eluting with the mAbs were common among the mAbs studied, with only a small percentage (∼10% on average) of a mAb's total HCP content in the protein A (PrA) eluate specific for a particular antibody. The abundance of these HCPs in cell culture fluids and their ability to interact with mAbs were the two main factors determining their prevalence in protein A eluates. Potential binding segments for HCPs to associate with mAbs were also studied through their co-purification with individual Fc and (Fab')2 antibody fragments. © 2016 American Institute of Chemical Engineers Biotechnol. Prog., 32:708-717, 2016.

Effect of high mannose glycan pairing on IgG antibody clearance.

IgG antibodies contain N-linked glycans on the Fc portion of each heavy chain. The glycan on one heavy chain can either match the glycan on the other heavy chain (symmetrical pairing) or be different (asymmetrical pairing). These Fc glycans influence effector functions and can alter clearance rates. Previous studies showing that high mannose forms result in faster mAb clearance in humans were incapable of differentiating the impact of symmetrically vs. asymmetrically paired HM forms, and, therefore, the effect of pairing on clearance was not clear. Traditional analytical methods, which are used to measure glycans in such studies, do not determine the number of HM glycans per antibody. With a sensitive method designed to measure HM pairing, we followed the levels of symmetrically and asymmetrically paired HM on antibodies in human pharmacokinetic serum samples to determine the impact of Fc HM glycan pairing on therapeutic human IgG clearance in humans. The two HM paired forms cleared at the same rate, indicating that the effect on clearance was not proportional to the degree of modification. Since both forms can exist on therapeutic antibodies and the ratio can differ between products, measuring their relative levels is necessary to properly estimate effects on clearance.

Development of advanced host cell protein enrichment and detection strategies to enable process relevant spike challenge studies.

An orthogonal chromatography methodology for the enrichment of host cell protein (HCP) species relative to monoclonal antibody (mAb) products was developed and applied for the successful enrichment of HCP from post-Protein A process pools for seven different mAb products. An advanced two-dimensional liquid chromatography/mass spectrometry platform (2D-LC/MS(E) ) was utilized to demonstrate that the HCP enriched material was representative, in terms of species content, to pre-enriched process pools. The HCP enrichment methodology was scaled up for two different mAb products, and this process relevant enriched HCP material was used to conduct advanced spike challenge studies to demonstrate the utility of the approach for the understanding of (1) quantitative HCP clearance, (2) individual species clearance, and (3) species clearance redundancy across polishing chromatography steps. The combined ability to enrich process relevant HCP, detect individual HCP species with 2D-LC/MS(E) technology, and conduct advanced challenge studies with process relevant material surmounts prior limitations to high integrity process challenge study implementation, and facilitates significant process understanding for development of risk-based control strategies and strategic process design. This also demonstrates implementation of a foundational strategy for conducting spike-challenge studies using process-relevant impurities isolated from processes of interest using orthogonal approaches.

Investigation of antibody disulfide reduction and re-oxidation and impact to biological activities.

Disulfide reduction in therapeutic monoclonal antibodies can occur during cell harvest operations as a result of cell breakage. Understanding these product quality changes and manufacturers' ability to control them would likely be of concern to regulatory bodies. To study the biological impact of disulfide reduction, mAbs, including IgG2κ, IgG2λ, IgG1κ, and IgG1λ forms, were partially reduced with dithiothreitol (DTT). Samples generated had approximately 10% or 50% intact molecules as determined by nrCE-SDS. Similar to the type of partial reduction obtained during uncontrolled harvest operations, DTT reduced antibodies were free from sulfur-linked adduct, such as attached cysteine. These partially reduced materials were incubated under physiological (blood-mimicking) redox conditions in vitro to follow the fate of the interchain cysteines. Within 8h, the original disulfide bonds reformed. For mAbA, an IgG2κ, the initial re-oxidized state favored the IgG2-A disulfide isoform, which then underwent conversion over time to other isoforms. Reduced material was fully active. Results suggest that the type of disulfide reduction would have minimal impact to safety or efficacy. Antibody re-oxidation rates were found to be in the order of IgG2κ

Impact of glycation on antibody clearance.

Glycation of therapeutic proteins occurs during mammalian cell culture expression and upon administration to patients. Since the chemical attachment of mannose or other sugars via a chemical linker has been shown to increase a protein's clearance rate in mice through the mannose receptor, we explored the effect of mannose glycation on the clearance of an IgG in mice. An IgG decorated with high levels of mannose (~18 mol/mol protein) through glycation did not clear faster in mice than the underivatized protein, whereas the same IgG decorated with mannose attached in a way to maintain the normal glycosidic bond (2-imino-2-methoxyethyl-1-thiomannoside, or IMT-mannose) at similar derivatization levels cleared significantly faster. Surface plasmon resonance studies revealed that the IgG derivatized with IMT-mannose bound tightly to the mannose receptor (KD = 20 nM) but the IgG glycated with mannose did not bind. These results indicate that glycation, even at unnaturally elevated levels, does not appear to be a clearance concern for therapeutic proteins.

Protected hinge in the immunoglobulin G2-A2 disulfide isoform.

Human IgG2 consists of disulfide-mediated structural isoforms, classified by the number of Fab arms disulfide-linked to the heavy chain hinge. In the IgG2-B isoform, both Fab arms are linked to the hinge region, and in IgG2-A, neither Fab arm are linked to the hinge. IgG2-A/B is a hybrid between these two forms, with only one Fab arm disulfide-linked to the hinge. Within each of these isoform types are subtypes, with subtle disulfide-linkage differences. Here we explored the structural basis for the A1 and A2 isoform subtypes. Whereas A1 isoform converts into the A/B and B isoforms under mild redox conditions, A2 does not. Characterization of the disulfide connectivities of A2 isoform revealed a similar structure to A1 isoform, with parallel inter heavy chain disulfide linkages in the hinge region. However, the hinge disulfides in A2 isoform were resistant to reduction under conditions where A1 isoform hinge disulfides became reduced and they required thermal treatment (>55 °C) to obtain thiol-dependent disulfide reduction. Structural analysis of the hinge region indicated that the protected disulfides were restricted to cysteines 219 and 220 of the upper hinge. Disruption of the upper hinge through insertion mutagenesis eliminated A2 isoform behavior. (1)H NMR studies showed that the A1 isoform Fc glycan was more dynamic than that on A2 isoform and showed some other conformational differences. Results point to an IgG2-A2 upper hinge region that is more akin to the interior of a globular protein than the flexible hinge region expected on an IgG.

Comprehensive tracking of host cell proteins during monoclonal antibody purifications using mass spectrometry.

An advanced two-dimensional liquid chromatography/mass spectrometry platform was used to quantify individual host cell proteins (HCPs) present at various purification steps for several therapeutic monoclonal antibodies (mAbs) produced in Chinese hamster ovary cells. The methodology produced reproducible identifications and quantifications among replicate analyses consistent with a previously documented individual limit of quantification of ~13 ppm. We were able to track individual HCPs from cell culture fluid to protein A eluate pool to subsequent viral inactivation pool and, in some cases, further downstream. Approximately 500 HCPs were confidently identified in cell culture fluid and this number declined progressively through the purification scheme until no HCPs could be confidently identified in polishing step cation-exchange eluate pools. The protein A eluate pool of nine different mAbs contained widely differing numbers, and total levels, of HCPs, yet the bulk of the total HCP content in each case consisted of a small subset of normally intracellular HCPs highly abundant in cell culture fluid. These observations hint that minimizing cell lysis during cell culture/harvest may be useful in minimizing downstream HCP content. Clusterin and actin are abundant in the protein A eluate pools of most mAbs studied. HCP profiling by this methodology can provide useful information to process developers and lead to the refinement of existing purification platforms.

Assessment of naturally occurring covalent and total dimer levels in human IgG1 and IgG2.

Antibody dimers, two self-associated monomers, have been detected on both recombinantly expressed and endogenous human IgG proteins. Nearly 10 years ago, Yoo et al. (2003) described low levels of IgG2 covalent dimer, in human serum, but did not quantify the levels. Here we quantify the total and covalent dimer levels of IgG2 and IgG1 in human blood, and study the origin of covalent dimer formation. Low levels (<1%) of total IgG1 and IgG2 dimers were measured in freshly prepared human plasma. Both IgG1 and IgG2 covalent dimers were also found in plasma. Whereas IgG1 covalent dimer levels were significantly reduced by steps intended to eliminate artifacts during sample preparation, IgG2 covalent dimer levels remain stable in such conditions. About 0.4% of IgG2 in plasma was in a covalent dimer form, yet very little (<0.03%) of IgG1 covalent dimer could be considered naturally occurring. IgG2 dimer also formed in vitro under conditions designed to mimic those in blood, suggesting that formation occurs in vivo during circulation. Thus, small amounts of covalent IgG2 dimer do appear to form naturally.

Cysteine racemization on IgG heavy and light chains.

Under basic pH conditions, the heavy chain 220-light chain 214 (H220-L214) disulfide bond, found in the flexible hinge region of an IgG1, can convert to a thioether. Similar conditions also result in racemization of the H220 cysteine. Here, we report that racemization occurs on both H220 and L214 on an IgG1 with a λ light chain (IgG1λ) but almost entirely on H220 of an IgGl with a κ light chain (IgG1κ) under similar conditions. Likewise, racemization was detected at significant levels on H220 and L214 on endogenous human IgG1λ but only at the H220 position on IgG1κ. Low but measurable levels of D-cysteines were found on IgG2 cysteines in the hinge region, both with monoclonal antibodies incubated under basic pH conditions and on antibodies isolated from human serum. A simplified reaction mechanism involving reversible ß-elimination on the cysteine is presented that accounts for both base-catalyzed racemization and thioether formation at the hinge disulfide.

Monoclonal antibody disulfide reduction during manufacturing: Untangling process effects from product effects.

Manufacturing-induced disulfide reduction has recently been reported for monoclonal human immunoglobulin gamma (IgG) antibodies, a widely used modality in the biopharmaceutical industry. This effect has been tied to components of the intracellular thioredoxin reduction system that are released upon cell breakage. Here, we describe the effect of process parameters and intrinsic molecule properties on the extent of reduction. Material taken from cell cultures at the end of production displayed large variations in the extent of antibody reduction between different products, including no reduction, when subjected to the same reduction-promoting harvest conditions. Additionally, in a reconstituted model in which process variables could be isolated from product properties, we found that antibody reduction was dependent on the cell line (clone) and cell culture process. A bench-scale model using a thioredoxin/thioredoxin reductase regeneration system revealed that reduction susceptibility depended on not only antibody class but also light chain type; the model further demonstrates that the trend in reducibility was identical to DTT reduction sensitivity following the order IgG1λ > IgG1κ > IgG2λ > IgG2κ. Thus, both product attributes and process parameters contribute to the extent of antibody reduction during production.

Profiling the effects of process changes on residual host cell proteins in biotherapeutics by mass spectrometry.

An advanced liquid chromatography/mass spectrometry (MS) platform was used to identify and quantify residual Escherichia coli host cell proteins (HCPs) in the drug substance (DS) of several peptibodies (Pbs). Significantly different HCP impurity profiles were observed among different biotherapeutic Pbs as well as one Pb purified via multiple processes. The results can be rationally interpreted in terms of differences among the purification processes, and demonstrate the power of this technique to sensitively monitor both the quantity and composition of residual HCPs in DS, where these may represent a safety risk to patients. The breadth of information obtained using MS is compared to traditional multiproduct enzyme-linked immunosorbent assay (ELISA) values for total HCP in the same samples and shows that, in this case, the ELISA failed to detect multiple HCPs. The HCP composition of two upstream samples was also analyzed and used to demonstrate that HCPs that carry through purification processes to be detectable in DS are not always among those that are the most abundant upstream. Compared to ELISA, we demonstrate that MS can provide a more comprehensive, and accurate, characterization of DS HCPs, thereby facilitating process development as well as more rationally assessing potential safety risks posed by individual, identified HCPs.

IgG1 thioether bond formation in vivo.

During either production or storage, the LC214-HC220 disulfide in therapeutic antibodies can convert to a thioether bond. Here we report that a thioether forms at the same position on antibodies in vivo. An IgG1κ therapeutic antibody dosed in humans formed a thioether at this position at a rate of about 0.1%/day while circulating in blood. Thioether modifications were also found at this position in endogenous antibodies isolated from healthy human subjects, at levels consistent with this conversion rate. For both endogenous antibodies and recombinant antibodies studied in vivo, thioether conversion rates were faster for IgG1 antibodies containing λ light chains than those containing κ light chains. These light chain reaction rate differences were replicated in vitro. Additional mechanistic studies showed that base-catalyzed thioether formation through the light chain dehydrogenation was more preferred on antibodies with λ light chains, which may help explain the observed reaction rate differences.

IgG2 disulfide isoform conversion kinetics.

Human IgG2 antibodies contain three types of disulfide isoforms, classified by the number of Fab arms having disulfide links to the heavy chain hinge region. In the IgG2-B form, both Fab arms have interchain disulfide bonds to the hinge region, and in IgG2-A, neither Fab arm are disulfide linked to the hinge. The IgG2-A/B is a hybrid between these two forms, with only one Fab arm disulfide linked to the hinge. Changes in the relative levels of these forms over time are observed while IgG2 circulates in humans, suggesting IgG2-A→IgG2-A/B→IgG2-B conversion. Using a flow-through dialysis system, we studied the conversion kinetics of these forms in vitro under physiological conditions. For two IgG2κ antibodies, in vivo results closely matched the kinetics observed in vitro, indicating that the changes observed in vivo were solely conversions between isoforms, not differential clearance of specific forms. Moreover, the combined results validate the accuracy of the physiological model for the study of blood redox reactions. Further exploration of the conversion kinetics using material enriched in the IgG2-A forms revealed that the IgG2-A→IgG2-A/B rate was similar between IgG2κ and IgG2λ antibodies. In IgG2κ antibodies, conversion of IgG2-A/B→IgG2-B was slower than the IgG2-A→IgG2-A/B reaction. However, in IgG2λ antibodies, little IgG2-A/B→IgG2-B conversion was detected under physiological conditions. Thus, small differences in the C-terminus of the light chain sequences affect the disulfide conversion kinetics and impact the IgG2 disulfide isoforms produced in vivo.

Evaluation of protein disulfide conversion in vitro using a continuous flow dialysis system.

Recombinant therapeutic proteins are heterogeneous due to chemical and physical modifications. Understanding the impact of these modifications on drug safety and efficacy is critical for optimal process development and for setting reasonable specification limits. In this study, we describe the development of an in vitro continuous flow dialysis system to evaluate potential in vivo behavior of thiol adducted species and incorrectly disulfide bonded species of therapeutic proteins. The system is capable of maintaining the low-level cysteine concentrations found in human blood. Liabilities of cysteamine adducted species, incorrectly disulfide bonded species, and the correctly disulfide bonded form of an Fc-fusion protein were studied using this system. Results showed that 90% of the cysteamine adduct converted into the correctly disulfide bonded form and incorrectly disulfide bonded species in approximately 4 h under physiological conditions. Approximately 50% of incorrectly disulfide bonded species converted into the correctly bonded form in 2 days. These results provide valuable information on potential in vivo stability of the cysteamine adduct, incorrectly disulfide bonded species, and the correctly bonded form of the Fc-fusion protein. These are important considerations when evaluating the criticality of product quality attributes.

Identification and quantification of host cell protein impurities in biotherapeutics using mass spectrometry.

Residual host cell proteins (HCPs) in biotherapeutics can present potential safety risks to patients or compromise product stability. As such, their levels are typically monitored using a multicomponent HCP enzyme-linked immunosorbent assay (ELISA) to ensure adequate removal. However, it is not possible to guarantee ELISA coverage of every possible HCP impurity, and the specific HCPs remaining following purification are rarely identified. In the current study, we characterized the ability of an advanced two-dimensional liquid chromatography/mass spectrometry platform (2D-LC/MS(E)) to identify and quantify known low-level spiked protein impurities in a therapeutic peptide Fc fusion protein. The label-free quantification procedure based on the "top 3" intensity tryptic peptides per protein was applied and improved on for this application. Limits of detection for unknown HCPs were approximated from the spiked protein data along with estimates for the quantitative accuracy of the method. In all, we established that most protein impurities present at 13±4ppm can be identified with a quantitative error of less than 2-fold using the more sensitive of two tested method formats. To conclude the study, we characterized all detectable Escherichia coli proteins present in this Fc fusion protein drug substance and discuss future applications of the method.

Rates and impact of human antibody glycation in vivo.

Glycation of immunoglobulin G (IgG) can result from incubation with a reducing sugar in vitro or during circulation in vivo. Upon injection of a recombinantly produced human therapeutic IgG into humans, changes in the glycation levels could be observed as a function of circulation time. Mass changes on the individual IgG polypeptide chains as the results of glycation were determined using reversed-phase liquid chromatography/mass spectrometry. Changes to the light and heavy chains were low but easily detectable at 0.00092 and 0.0021 glucose (Glc) additions per chain per day, respectively. Levels of glycation found on the Fc portion of IgG isolated from healthy subjects, using a similar analytical approach, were on average 0.045 Glc molecules per fragment. In vivo glycation rates could be approximated in vitro by modeling the physiological glycation reaction with a simplified incubation containing physiological Glc concentrations, pH and temperature but with a high concentration of a single purified IgG. To test the impact of glycation on IgG function, highly glycated IgG1 and IgG2 were prepared containing on average 42-49 Glc molecules per IgG. Binding to FcγIIIa receptors, neonatal Fc receptor or protein A was similar or identical to the non-glycated IgG controls. Although the modifications were well distributed throughout the protein sequence, and at high enough levels to affect the elution position by size-exclusion chromatography, no changes in the tested Fc functions were observed.

Rapid LC-MS screening for IgG Fc modifications and allelic variants in blood.

A new method for simultaneously screening allelic variants and certain Fc modifications on endogenous human IgG1 and IgG2 directly from blood samples is described. The IdeS endoproteinase was used to cleave IgG in serum to generate Fc, which, after denaturation, was analyzed directly as monomeric Fc (Fc/2) by LC-MS to identify the haplotype(s) present in each individual. The relative levels of IgG isotype and haplotype ratios were generated along with the profile of the major Fc glycans and several other modifications associated with each IgG1 or IgG2 haplotype. Since only minute quantities (5 µL) of blood are required and analysis can be highly automated, this approach lends itself to screening large populations. We demonstrate its utility in examining possible correlations between Fc properties and allelic variants. IgG1 core fucosylation, which significantly impacts antibody dependent cellular cytotoxicity (ADCC), showed an asymmetric distribution, with a small number of individuals showing unexpectedly high core afucosylation levels. In these individuals, IgG2 afucosylation levels were normal. Finally, a new IgG1 allotype, previously not characterized, was identified using this analytical methodology.