Influence of N-glycosylation on effector functions and thermal stability of glycoengineered IgG1 monoclonal antibody with homogeneous glycoforms

Abstract

Glycosylation of the conserved asparagine residue in each heavy chain of IgG in the CH2 domain is known as N-glycosylation. It is one of the most common post-translational modifications and important critical quality attributes of monoclonal antibody (mAb) therapeutics. Various studies have demonstrated the effects of the Fc N-glycosylation on safety, Fc effector functions, and pharmacokinetics, both dependent and independent of neonatal Fc receptor (FcRn) pathway. However, separation of various glycoforms to investigate the biological and functional relevance of glycosylation is a major challenge, and existing studies often discuss the overall impact of N-glycans, without considering the individual contributions of each glycoform when evaluating mAbs with highly heterogeneous distributions. In this study, chemoenzymatic glycoengineering incorporating an endo-β-N-acetylglucosaminidase (ENGase) EndoS2 and its mutant with transglycosylation activity was used to generate mAb glycoforms with highly homogeneous and well-defined N-glycans to better understand and precisely evaluate the effect of each N-glycan structure on Fc effector functions and protein stability. We demonstrated that the core fucosylation, non-reducing terminal galactosylation, sialylation, and mannosylation of IgG1 mAb N-glycans impact not only on FcγRIIIa binding, antibody-dependent cell-mediated cytotoxicity, and C1q binding, but also FcRn binding, thermal stability and propensity for protein aggregation.

Keywords: Fc effector functions; Fc glycosylation; Fc receptor; IgG1; Monoclonal antibody (mAb); antibody-dependent cell-mediated cytotoxicity (ADCC); chemoenzymatic transglycosylation; complement dependent cytotoxicity (CDC); protein aggregation; thermal stability.

Figure 1.

Generation of IgG1 glycoforms with homogeneous glycoforms. (a) IgG1 N-glycosylation schematics and abbreviations used in this study for elucidating structures of complex type biantennary N-glycan. Presence of a core fucose is indicated by F; Ax indicates a biantennary glycan structure with the number (x) of GlcNAc on the antenna; Gx indicates the number (x) of galactose on the antenna; Sx indicates the number (x) of sialic acid on the antenna. Fuc: Fucose, Man: Mannose, GlcNAc: N-Acetylglucosamine, Gal: Galactose, SA: Sialic Acid. (b) Each core fucosylated and non-fucosylated homogeneous glycoform was generated from the same CHO cell line with or without Kifunensine to produce starting mAbs, control (ctrl) and high-mannose (HM). Subsequent deglycosylation and transglycosylation performed to obtain resulting homogeneous glycoforms. Endoglycosidase EndoS2 from S. pyogenes and D184 mutant as glycosynthase recombinantly expressed from E. coli. Glycan oxazolines (oxa) were used as well-defined donor substrates (Figure S1).

Figure 2.

Reduced Heavy chain LC-MS analysis of glycoforms. Deconvoluted mass spectrometric data of heavy chain of starting mAbs, acceptors and resulting six glycoforms are shown to confirm intended transglycosylation and good agreement with theoretical molecular mass shown in Table 2. Starting mAbs showed relatively heterogeneous mass spectrum, where resulting acceptors and transglycosylated glycoforms showed highly homogeneous profile. Truncation of one sialic acid residue observed for mAb-FA2G2S2 and mAb-A2G2S2 is possibly due to fragmentation induced by ionization. No alternations in light chain were observed by the reduced mass analysis (data not shown).

Figure 3.

2-AB labeled released N-glycan analysis of the transglycosylated glycoforms. N-glycans of heavy chain of the transglycosylated glycoforms were enzymatically released by PNGaseF and labelled by 2-AB for fluorescent detection in the HILIC mode separation to confirm the intended enrichment of N-glycans. Relative peak abundance of each glycoform was also described in Table 1, showing over 90% homogeneity for all transglycosylated glycoforms generated.

Figure 4.

Binding affinity of FcγRIIIa was increased by core defucosylation and terminal galactosylation, decreased by further terminal sialylation in SPR analysis. (a) Sensorgrams of each glycoform. Anti-His Ab was immobilized on CM5 chip. N-terminal his tagged FcγRIIIa captured and glycoforms as analytes were injected by single-cycle mode at five concentrations. (b) Relative binding activities of FcγRIIIa. Error bars represent SD of n = 3 and asterisks indicate statistical significance at p < 0.05. n.d: not determined. (c) JMP analysis: Scaled estimates of effect of glycosylation for binding affinity of FcγRIIIa. Analyzed for complex biantennary glycoforms. (d) JMP analysis: Interaction profile of terminal glycosylation for binding affinity of FcγRIIIa. Analyzed for complex biantennary glycoforms. In lower left panel, red: Gal, blue: SA, and green: GlcNAc. Significant statistical interaction between core fucosylation and terminal galactosylation (c-d).

Figure 5.

FcγRIIIa affinity chromatography showed longer retention by core fucosylation, terminal galactosylation and further sialylation. Non-glycosylated recombinant FcγRIIIa prepacked column was used to analyze the affinity interaction between transglycosylated glycoforms.

Figure 6.

ADCC activity was increased by core defucosylation and terminal galactosylation, decreased by further sialylation in reporter bioassay for FcγRIIIa. Nuclear factor of activated T-cells (NFAT) engineered Jurkat cells expressing FcγRIIIa were incubated with target antigen expressing CHO-K1 cells. Receptor binding activates gene transcription through the NFAT pathway in effector cell leading to luciferase production and subsequent luciferase substrate addition was followed by luminescence detection. (a) Response curves of mAb-ctrl and fucosylated glycoforms. (b) Response curves of mAb-ctrl and non-fucosylated glycoforms. (c) Relative binding activity from EC₅₀ of glycoforms. n.d: not determined. (d) JMP analysis: scaled estimates of effect of glycosylation for ADCC activity. Analyzed for complex biantennary glycoforms. (e) JMP analysis: Interaction profile of terminal glycosylation for ADCC activity. Analyzed for complex biantennary glycoforms. In lower left panel, red: Gal, blue: SA, and green: GlcNAc. Significant statistical interaction between core fucosylation and terminal galactosylation (d-e). Error bars represent 95% confidence interval of n = 3 (a-c).

Figure 7.

Binding affinity of C1q increased by terminal galactosylation and sialylation. (a) Sensorgrams of each glycoform. Glycoforms were first captured on Protein A chip and C1q as analytes were injected by multi-cycle mode at eight concentrations. Sensorgrams were fitted with steady-state equilibrium model. (b) Relative binding activities of C1q. Error bars represent SD of n = 3 and asterisks indicate statistical significance at p < 0.05. (c) JMP analysis: Scaled estimates of effect of glycosylation for binding affinity of C1q. Analyzed for complex biantennary glycoforms and acceptors.

Figure 8.

No effect of glycosylation for binding affinity of FcRn in SPR analysis. (a) Sensorgrams of each glycoform. FcRn was directly immobilized on CM5 chip and glycoforms as analytes were injected by multi-cycle mode at six concentrations. Sensorgrams were fitted with steady-state equilibrium model. (b) Relative binding activities of FcRn. Error bars represent variation range of n = 2.

Figure 9.

FcRn affinity chromatography showed longer retention by terminal galactosylation and sialylation. FcRn affinity chromatography utilized a human FcRn prepacked column. (a) Chromatograms and (b) Retention time of main peak. Terminal galactosylation and sialylation resulted in longer retention where mAb-HM and partially deglycosylated glycoforms showed the shortened retention. Error bar for mAb-ctrl represents SD of n = 6.

Figure 10.

Differential scanning calorimetry analysis of glycoforms showed thermal stability of CH2 domain was increased by terminal galactosylation and decreased in partially deglycosylated acceptors, terminally mannosylated mAb-HM, core defucosylated and terminal sialylated glycoforms. (a) Thermograms of each glycoform. (b) Transition peak of CH2 domain obtained from non-two-state curve fitting. (c) Thermal transition temperature T_m of CH2 domain obtained from curve fitting. T_m CH2 was significantly reduced in partially deglycosylated acceptors and terminally mannosylated mAb-HM. (d) Denaturation enthalpy ΔH of CH2 domain obtained from curve fitting showed significantly increased denaturing enthalpy by terminal galactosylation and decreased by core defucosylation and terminal sialylation. Error bars were generated based on variance of curve fitting using Origin software (c-d).

Figure 11.

Intrinsic fluorescence spectroscopy (FS) analysis to measure transition temperature showed transition temperature was decreased in partially deglycosylated acceptors and terminally mannosylated mAb-HM glycoforms. (a) Barycentric mean (BCM) shift curve. (b) First derivative of BCM shift and (c) peak top temperatures as T_m FS for each glycoform. T_m FS was significantly reduced in partially deglycosylated acceptors and terminally mannosylated mAb-HM. Error bars represent variation range of n = 2.

Figure 12.

Static light scattering analysis to measure aggregation propensity by temperature ramping showed aggregation temperature was increased by terminal galactosylation and decreased by terminal sialylation. (a) SLS counts of each glycoform. (b) Aggregation temperatures (T_agg) showed increased stability by terminal galactosylation and decreased stability by terminal sialylation. Core defucosylation also showed slightly decreased T_agg. Error bars represent variation range of n = 2.

Figure 13.

Static light scattering analysis to measure aggregation propensity by isothermal incubation at 65 °C showed aggregation speed was decreased by terminal galactosylation and increased by core defucosylation and terminal sialylation. (a) SLS counts for seven hours. (b) Time required to reach 2, 10 and 50% of maximum SLS counts for each glycoform. Error bars represent SD of n = 3. (c) JMP analysis: Scaled estimates of effect of glycosylation for relative time required to reach certain levels. Time required to reach 2, 10 and 50% of maximum SLS counts was converted into relative time to average of mAb-ctrl and statistically analyzed for complex biantennary glycoforms and acceptors.

Figure 14.

Long-term forced degradation analyzed by size-exclusion chromatography showed significant increase of aggregation species in acceptors, mAb-HM and terminally galactosylated glycoforms but overall the same level of fragmentation among all glycoforms tested. (a) Chromatographic comparison of glycoforms after incubation for 17 days. (b) % increase of HMW species, with each aggregate-1 and aggregate-2 on top and bottom, respectively. (c) % increase of LMW species, with each fragment-1 and fragment-2 on bottom and top, respectively. n.t: not tested (b-c).