Isolation of Monoclonal Antibodies Using Capillary Electrophoresis

Antibodies or immunoglobulins are soluble serum glycoproteins that offer passive immunity against foreign antigens. Monoclonal Antibodies (MAbs) have been developed as therapeutic reagents because they can specifically treat particular molecular targets associated with disease manifestation. There is a high degree of functional and structural heterogeneity among antibodies, especially because of the diversity of associated glycosylation. Glycosylation on monoclonal therapeutic antibodies is a critical post-translational modification that is linked with their structure, bioactivity and pharmacokinetics. A number of different carbohydrate moieties can potentially bind to MAbs, but it is generally understood that a core group of bi-antennary and high-mannose structures make up the most commonly associated species. MAb carbohydrate heterogeneity analysis and quantitation is important as oligosaccharides linked to their Fc region play a significant role in the regulation of Complement Dependent Cytotoxicity (CDC) and Antibody Dependent Cellular Cytotoxicity (ADCC).

Glycan species with a variation in terminal Gal content can be separated and analyzed readily using existing CE technology. Glycan sample preparation includes addition of both charge and fluorescence properties enabling oligosaccharides to be electrophoretically separated and then quantified using laser- induced fluorescence (LIF) detection technology.

Procedure of Glycan Analysis Sample Preparation

The procedure of glycan analysis sample preparation is listed below:

  • First, oligosaccharides are separated from the Asn297 residue of the MAb backbone using the N-glycosidase F (PNGase F).
  • Next, the fluorophore 8-aminopyrene-1, 3, 6-tri-sulfonic acid (APTS) is derivatized via reductive amination at the reducing end of the oligosaccharide as shown in figure 1.
  • Electrophoretic separation is carried out using a polymeric separation matrix comprising 0.4% polyethylene oxide (PEO). Beckman Coulter has designed and commercialized the technology in order to automate and simplify this process. It has been shown that the principle for this gel-based CE separation of oligosaccharides is based on both mobility and hydrodynamic volume. This is shown in part by the fact that positional isomers, although identical in mass, can be resolved from one another as shown in figure 2.
  • Modifications on glycan structures including the presence of fucose, terminal sialic acid, or a bisecting N-Acetylglucosamine (GlcNAc) have been associated with changes in ADCC activity, thus affecting MAb efficacy.
  • As the size difference between fucosylated and afucosylated glycans is as small as 16 daltons, and that they may have numerous positional isomers, separation has proven to be difficult for certain species.

Schematics of glycan analysis sample preparation and various carbohydrate structures. A. Glycan cleavage and APTS derivatization strategy for analysis of N-linked oligosaccharides. B. Examples of 2 glycan species: N-linked oligosaccharide illustrating putative important modifications (left) and high mannose structure (right).

Figure 1. Schematics of glycan analysis sample preparation and various carbohydrate structures. A. Glycan cleavage and APTS derivatization strategy for analysis of N-linked oligosaccharides. B. Examples of 2 glycan species: N-linked oligosaccharide illustrating putative important modifications (left) and high mannose structure (right).

Current methods have not been able to resolve a large number of the major co-migrating glycan species from each another. Previously, CE conditions capable of separating fucosylated and afucosylated N-linked oligosaccharides were presented.

Along with the success in CE technology, they help resolving differences in terminal galactosylation, suggesting that it should also be possible to separate glycans that are fucosylated, sialylated or bisected from each other as well as high mannose species.

Methods and Materials

In the case study analyzed in this article, all separations were carried out using the PA 800 plus Pharmaceutical Analysis System configured with a 488-nm solid state laser and LIF detection with an emission band-pass filter of (520 ± 10) nm. N-CHO capillaries were used for separation of ligosaccharides. All other assay conditions were as described in the standard operating procedure for the Carbohydrate Labeling and Analysis Assay Kit with the exception that carbohydrate separation buffer was replaced with a new separation buffer formulation, where indicated. The final concentration for oligosaccharide samples was 1.25 μM. Glycan standards for fucosylated and afucosylated species of G0, G1, G1’ and G2 were bought from Glyko ProZyme (Hayward, CA).The therapeutic MAb was obtained from Genentech (San Francisco, CA).

The experimental details of this work are listed below:

  • Carbohydrate separation gels used:
    • Carbohydrate assay gel (contains polyethylene oxide (PEO)).
    • New separation gel buffer was 1:1 mixture of:
      • Carbohydrate separation gel buffer (PEO) – BEC p/n 477623.
      • dsDNA1000 separation gel buffer (LPA) – BEC p/n 477628.
  • Capillary length: total length = 60.2 cm, length to detector = 50 cm .
  • Capillary diameter: 50 μm I.D.
  • Injection conditions: 0.5 psi for 10 sec unless otherwise stated.
  • Separation Voltage: 30 kV.
  • Field Strength: 500 volts/cm.
  • Capillary cartridge temperature: 20° C.
  • Sample storage temperature: 10° C.

Results and Discussion

The objective of this study was to achieve separation of major complex glycan species associated with monoclonal antibodies. This glycan population includes oligosaccharides with and without core fucose moieties, terminal galactose subunits, terminal sialic acids and bisecting GlcNAc residues as well as numerous positional isomers. The separation limitations for existing separation chemistry and the Beckman Coulter Carbohydrate Labeling & Analysis Assay Kit were characterized. Employing standard protocols which is included with the Carbohydrate Labeling & Analysis Assay for instrument configuration, sample preparation and separation conditions, it was possible to obtain baseline resolution between G0, G1 positional isomers (G1 and G1’) and G2 oligosaccharide species (Fig. 2).

Separation of G0, G1, and G2 glycan species. Representative data (top trace) shows separation of N-linked oligosaccharides G0, G1, G1’ and G2 using the Carbohydrate Labeling & Analysis Assay Kit (Beckman Coulter p/n 477600).

Minutes

Figure 2. Separation of G0, G1, and G2 glycan species. Representative data (top trace) shows separation of N-linked oligosaccharides G0, G1, G1’ and G2 using the Carbohydrate Labeling & Analysis Assay Kit (Beckman Coulter p/n 477600).

A systematic approach was formulated in which standards were spiked into samples to help identify additional peaks in this separation and also to better define co-migration of glycans that may be occurring. The G0, G1 and G2 species are shown in Figure 3.

Optimization of the carbohydrate separation buffer allows for resolution between closely-migrating oligosaccharide pairs. Using standard sample preparation protocols, oligosaccharide standards were APTS labelled and separated by CE. Resolution of co-migrating was facilitated by combining existing Beckman Coulter separation buffers. Separation buffer consisted of a 1:1 mixture of Carbohydrate Separation Buffer (Beckman Coulter p/n 477623) containing 0.4% PEO and dsDNA 1000 Gel Buffer (Beckman Coulter p/n 477628) containing a low percentage of linear polyacrylamide (LPA). Separations were performed on an N-CHO capillary (p/n 477600) with an effective length of 50 cm. Separation conditions were 20 kV following 0.5 psi injection for 5 seconds. Field strength was 333 V/cm. (A) Closely migrating fucosylated and afucosylated N-linked oligosaccharide standards G0F and G1 (red labels) as well as G1’F and G2 (blue labels) were separated from one another. (B) Closely migrating high mannose oligosaccharide standards were separated from one another. This separation resolved Man-5 from G0, Man-6 from G0F, and Man-7 from both G1F and G2.

Minutes

Optimization of the carbohydrate separation buffer allows for resolution between closely-migrating oligosaccharide pairs. Using standard sample preparation protocols, oligosaccharide standards were APTS labelled and separated by CE. Resolution of co-migrating was facilitated by combining existing Beckman Coulter separation buffers. Separation buffer consisted of a 1:1 mixture of Carbohydrate Separation Buffer (Beckman Coulter p/n 477623) containing 0.4% PEO and dsDNA 1000 Gel Buffer (Beckman Coulter p/n 477628) containing a low percentage of linear polyacrylamide (LPA). Separations were performed on an N-CHO capillary (p/n 477600) with an effective length of 50 cm. Separation conditions were 20 kV following 0.5 psi injection for 5 seconds. Field strength was 333 V/cm. (A) Closely migrating fucosylated and afucosylated N-linked oligosaccharide standards G0F and G1 (red labels) as well as G1’F and G2 (blue labels) were separated from one another. (B) Closely migrating high mannose oligosaccharide standards were separated from one another. This separation resolved Man-5 from G0, Man-6 from G0F, and Man-7 from both G1F and G2.

Minutes

Figure 3. Optimization of the carbohydrate separation buffer allows for resolution between closely-migrating oligosaccharide pairs. Using standard sample preparation protocols, oligosaccharide standards were APTS labelled and separated by CE. Resolution of co-migrating was facilitated by combining existing Beckman Coulter separation buffers. Separation buffer consisted of a 1:1 mixture of Carbohydrate Separation Buffer (Beckman Coulter p/n 477623) containing 0.4% PEO and dsDNA 1000 Gel Buffer (Beckman Coulter p/n 477628) containing a low percentage of linear polyacrylamide (LPA). Separations were performed on an N-CHO capillary (p/n 477600) with an effective length of 50 cm. Separation conditions were 20 kV following 0.5 psi injection for 5 seconds. Field strength was 333 V/cm. (A) Closely migrating fucosylated and afucosylated N-linked oligosaccharide standards G0F and G1 (red labels) as well as G1’F and G2 (blue labels) were separated from one another. (B) Closely migrating high mannose oligosaccharide standards were separated from one another. This separation resolved Man-5 from G0, Man-6 from G0F, and Man-7 from both G1F and G2.

Spiking experiments using oligosaccharide standards illustrated that individual separated peaks may contain multiple glycan species. This was demonstrated before by co-migration of G0+fucose (GoF) with G1 and co-migration of G1’+fucose (G1’F) with G2. Modification of separation parameters that include capillary length, separation voltage, and temperature did not offer improved resolution.

By developing a new separation buffer formulation, it was possible to better resolve these co-migrating species (Fig. 3). Additional spiking experiments showed the power of the CE separation developed (Fig. 4). For testing this separation method on an antibody, a therapeutic MAb was obtained and its associated glycans (Fig. 5) were analyzed. Spiking with oligosaccharide standards to help identify glycan species resulted in good resolution between many of the major oligosaccharides, which were previously difficult to separate by CE.

Table 1 indicates the compound names, abbreviations used in the data, as well as descriptions and molecular weight of each glycan species.

Table 1. Glycan abbreviations and descriptions. In the course of this work, separation of standards was utilized to help identify various glycan peak positions.

Abbreviation Description Compound Name MW (Da)
-G Trimannosyl core M3N2 911
-GF Trimannosyl core, substituted with fucose M3N2F 1057
Man-5 Oligomannose 5 Man-5 1235
G0 Asialo, agalacto, biantennary complex NGA2 1317
Man-6 Oligomannose 6 Man-6 1398
G0F Asialo, agalacto, biantennary complex, core substituted with fucose NGA2F 1463
G1 / G1’ Asialo, monogalactosylated, biantennary complex NA2G1 1480
Man-7 Oligomannose 7 Man-7 1560
G1F / G1’F Asialo, mono-galactosylated, biantennary complex, core substituted with fucose NA2G1F 1626
G2 Asialo, galactosylated, biantennary complex NA2 1641
G0FB Asialo-, agalacto-, biantennary, core-substituted with fucose and bisecting N-acetylglucosamine (GlcNAc) NGA2FB 1667
Man-8 Oligomannose 8 Man-8 1722
G2F Asialo, galactosylated, biantennary complex, core-substituted with fucose NA2F 1787
Man-9 Oligomannose 9 Man-9 1884
G2S1 Mono-sialylate, galactosylated, biantennary complex A1 1933
G2S1F Mono-sialylate, galactosylated, biantennary complex, core-substituted with fucose A1F 2079

Optimization of the carbohydrate separation buffer allows for resolution of a number of MAb-associated oligosaccharides. Using standard sample preparation protocols, oligosaccharide standards were APTS labeled and separated by CE. Resolution of N-linked oligosaccharide standards was facilitated by combining existing Beckman Coulter separation buffers.

Minutes

Figure 4. Optimization of the carbohydrate separation buffer allows for resolution of a number of MAb-associated oligosaccharides. Using standard sample preparation protocols, oligosaccharide standards were APTS labeled and separated by CE. Resolution of N-linked oligosaccharide standards was facilitated by combining existing Beckman Coulter separation buffers.

Separation of oligosaccharides associated with a recombinant therapeutic MAb. Oligosaccharides were cleaved from a therapeutic MAb, APTS labeled and separated by CE using the new buffer formulation. A number of oligosaccharide species associated with this MAb were resolved from one another (A). In order to identify and help illustrate resolution between co-migrating glycan species, the MAb sample was spiked with standards (B). Relative to the oligosaccharide standards, it was possible to quantify G0, Man-5, G0F, G0FB, Man-7, G1F, G1’F and G2F. G2 standard (in blue) was also spiked into the mixture to indicate the location in the separation at which this oligosaccharide species would reside.

Minutes

Figure 5. Separation of oligosaccharides associated with a recombinant therapeutic MAb. Oligosaccharides were cleaved from a therapeutic MAb, APTS labeled and separated by CE using the new buffer formulation. A number of oligosaccharide species associated with this MAb were resolved from one another (A). In order to identify and help illustrate resolution between co-migrating glycan species, the MAb sample was spiked with standards (B). Relative to the oligosaccharide standards, it was possible to quantify G0, Man-5, G0F, G0FB, Man-7, G1F, G1’F and G2F. G2 standard (in blue) was also spiked into the mixture to indicate the location in the separation at which this oligosaccharide species would reside.

Conclusion

High resolution CE separations based on mobility and hydrodynamic volume have been developed for quantitative analysis of glycans. Using published protocols and commercially available reagents, it has been proved that this technology can separate oligosaccharides differing in terminal galactose. It has also been shown that by combining standard PEO separation gel buffer with a LPA gel buffer, it is possible to separate fucosylated from afucosylated N-linked oligosaccharides, high mannose structures, and numerous other glycan moieties. This work indicates that CE can be used to successfully separate and quantify a wide array of N-linked oligosaccharides associated with MAbs.

This information has been sourced, reviewed and adapted from materials provided by Beckman Coulter, Inc. - Particle Characterization.

For more information on this source, please visit Beckman Coulter, Inc. - Particle Size Characterization.

Ask A Question

Do you have a question you'd like to ask regarding this article?

Leave your feedback
Submit