Data on charge separation of bispecific and mispaired IgGs using native charge-variant mass spectrometry

The data supplied in this work are related to the research article entitled “Characterization of Bispecific and Mispaired IgGs by Native Charge-Variant Mass Spectrometry” (Phung et al., 2019). This data article describes a powerful analytical platform using native weak cation exchange chromatography coupled to a high-resolution mass spectrometer, charge variant mass spectrometry (CV-MS), to characterize bispecific and mispaired antibody species. Elution order is investigated through analytical methods and molecular modeling in an effort to understand the intrinsic charge, size and shape differences of these molecules.


a b s t r a c t
The data supplied in this work are related to the research article entitled "Characterization of Bispecific and Mispaired IgGs by Native Charge-Variant Mass Spectrometry" (Phung et al., 2019). This data article describes a powerful analytical platform using native weak cation exchange chromatography coupled to a high-resolution mass spectrometer, charge variant mass spectrometry (CV-MS), to characterize bispecific and mispaired antibody species. Elution order is investigated through analytical methods and molecular modeling in an effort to understand the intrinsic charge, size and shape differences of these molecules.

Value of the data
• Charge variant mass spectrometry (CV-MS) was utilized to successfully separate the correctly paired and light chain-scrambled mispaired isobaric species which may occur during bispecific antibody production in single host cells. • Scientific researchers in a variety of settings can utilize CV-MS to separate proteoforms based on only minor charge differences and directly analyze by mass spectrometry. • By charge separation of correctly paired and mispaired species, downstream purification of BsIgGs are made possible as an alternative to the incorporation of chain pairing strategies. • Molecular modeling revealed that minor surface exposed charge patches created microenvironments which were sufficient to resolve species, but not significant enough to alter the molecular isoelectric point.

Data
Our data demonstrates an integrated native ion exchange chromatography mass spectrometry based analytical method, CV-MS [1 , 2] , to successfully separate correctly paired and mispaired IgG species which may arise during bispecific antibody production in single host cells [1] .
Anti-HER2/CD3 BsIgG was analyzed by reversed phase HPLC using an organic solvent gradient ( Fig. 1 ) and traditional ion exchange using a salt gradient ( Fig. 2 (a)). We were not able to resolve isobaric species with extensive method optimization. A commercially available pH buffer system (CX-1) was also used to confirm the separation of IgG species with pH gradient cation chromatography ( Fig. 2 (b)), but these buffers are incompatible with downstream mass spectral analysis. Charge variant-mass spectrometry (CV-MS) was the only method demonstrated to resolve BsIg scrambled species, and holds the additional benefit of mass spectral detection for species identification [1] . To demonstrate the separation of impurities in a variety of paired single cell half antibody IgG assemblies by CV-MS, anti-IL-13/IL-4 and anti-EGFR/MET were also analyzed after optimizing the pH gradient for each BsIgG ( Fig. 3 ) and show complete separation of the main peak from the BsIg scramble.
To confirm the identity of isobaric species in the single cell anti-HER2/CD3 BsIg, each peak was fraction collected after charge separation. The isoelectric point was compared for each IgG species by iCIEF using iCE3 ( Fig. 4 ). The hydrodynamic radius was also compared for each IgG species by SEC-MALS ( Fig. 5 ). Lastly, molecular modeling with BioLuminate was used to generate protein titration curves of charge over pH for anti-HER2/CD3 in the research article [1] , as well as two other molecules, anti-IL-13/IL-4 and anti-EGFR/MET ( Figs. 6 and 7 ). Molecular modeling was employed to localize charge changes for each IgG species over a pH range from 7 to 9 and the surface exposed charge patches were compared ( Figs. 8 and 9 ).

Experimental design, materials, and methods
Below is a brief description of the experimental methods used to acquire data in this paper. For a more detailed and thorough report, please refer to the related research article [1] .

Native CV-MS
Bispecific antibodies were buffer exchanged into 50 mM ammonium acetate and separated on a ProPac WCX-10 column (Thermo Scientific) using gradients optimized for each individual BsIgG. The HPLC was coupled to a Thermo Exactive Plus EMR Orbitrap (Thermo Fisher Scientific). Acquired data was processed with Thermo Xcalibur Qual Brower and Thermo Protein Deconvolution 4.0.

Offline reversed phase separation and ion exchange
Separation of anti-HER2/CD3 BsIgG was evaluated on a HALO RP column and analyzed on a Thermo Exactive Plus EMR Orbitrap instrument for native CV-MS comparison. Anti-HER2/CD3 BsIgG was also buffer exchanged into 50 mM ammonium acetate, pH 7 and separated offline   using a ProPac WCX-10 column on a Shimadzu FRC-10A instrument. Traditional IEC salt gradient and commercial pH gradient (CX-1 pH Gradient Buffer Kit, Thermo Scientific), which are incompatible with downstream mass spectral analysis, were used for offline separation and compared to native CV-MS.

Comparison of anti-HER2/CD3 bispecific and mispaired IgGs by iCIEF and SEC-MALS
Buffer exchanged anti-HER2/CD3 BsIgG was separated and fractionated offline using a ProPac WCX-10 column on a Shimadzu FRC-10A instrument. Fractionated anti-HER2/CD3 bispecific and mispaired IgGs were analyzed by iCIEF using iCE3 to determine differences in isoelectric point. Fractionated samples were also analyzed by SEC-MALS using quasi elastic light scattering (QELS) to determine differences in hydrodynamic radius. The Stokes-Einstein relationship was used to calculate R H from measured diffusion coefficients.