Optimization of a reversed-phase high-performance liquid chromatography/mass spectrometry method for characterizing recombinant antibody heterogeneity and stability

https://doi.org/10.1016/j.chroma.2006.01.016Get rights and content

Abstract

An enhanced analytical RP-HPLC/MS method was developed for monitoring the stability and production of intact and fragmented monoclonal antibodies (MAbs). The use of high column temperatures (70–80 °C), organic solvents with high eluotropic strength coefficients (isopropyl and n-propyl alcohols), and Zorbax StableBond columns, were critical for good recovery and resolution of immunoglobulin G1 (IgG1) and IgG2 monoclonal antibodies. Using this method, cleavage products of a degraded IgG1 antibody were clearly separated and identified by in-line electrospray ionization time-of-flight (ESI-TOF) mass spectrometry generating exact masses and unique terminal ladder sequences. The glycosylation profile, including mapping of the terminal galactose and fucose heterogeneity of the N-linked sugars, was determined by mass spectrometry of intact MAbs. In addition, we discovered that several IgG2 MAbs exhibited greater structural heterogeneity compared to IgG1s. Mass spectral characterization data and reduction data suggested that the heterogeneity is disulfide related. This reversed-phase LC/MS method represents a key advancement in monitoring intact MAb production and stability.

Introduction

Reversed-phase high-performance liquid chromatography (RP-HPLC) analysis is a widely used analytical technique for monitoring the stability and production of biomolecules. The ability of this method to resolve nearly any form of chemical modification at a peptide and/or protein level and the direct adaptability to in-line mass spectrometry has proven to be a powerful tool. However, the use of RP-HPLC as an analytical tool for monitoring intact monoclonal antibodies (MAbs) has been limited because of the complex and hydrophobic nature of these large macromolecules causing poor recovery and limited resolution.

Although peptide mapping can be used to obtain detailed characterization of high molecular weight proteins, this method is time-consuming for sample preparation and data interpretation and becomes very complicated for large proteins. There have been examples of punitive modification caused by the digest conditions that are not easily recognized [1]. In addition, post-translational modifications can confound their analyses in mass spectrometry and/or HPLC. Thus, despite the fact that there are techniques that have been extensively used in the analysis of low molecular weight proteins or low molecular weight digests of larger proteins, there remains a need for additional methods and techniques for producing sequence and conformational information about intact proteins of high molecular weight.

Previous efforts to develop stability-indicating RP-HPLC methods for the analysis of high molecular weight proteins, including antibodies, have had limited success. Reversed-phase chromatography using Poros columns has been the standard technique for analyzing intact and reduced antibodies by LC/MS [2], [3], [4], [5], [6]. This technique has been shown to produce narrow peaks but suffers from the following two limitations. First, a relatively small value of selectivity factor (a), determined by the low surface density of Poros material, limits separation of possible structural variants and some degradation products with similar retention properties. Second, the high flow rate required for efficient separation on Poros columns reduces sensitivity of in-line mass spectrometric detection. The electrospray ionization mass spectrometers behave as concentration sensitive detectors and greatly benefit from the elution of protein fraction with low flow rates [7].

Recent advancements in method development have shown promise in improving reversed-phase chromatography of MAbs by using the long alkyl chains of the stationary phase, column temperatures elevated to 65–70 °C, and combination of trifluoroacetic acid (TFA) and heptafluorobutyric acid (HFBA) ion-pairing agents [8]. However, these efforts in RP-HPLC of antibodies have not been successful for some IgG subtypes and other more hydrophobic MAbs. For enhanced chromatography of these antibodies, further optimization was required, particularly in terms of the solvent composition of the mobile phase and column temperature.

Section snippets

Materials

Recombinant monoclonal IgG antibodies were expressed in CHO cells and produced at Amgen Inc. (Thousand Oaks, CA, USA). The purified proteins were stored in formulation buffer at 4 or −70 °C. Other commercially available recombinant monoclonal antibodies were tested for comparison. Dithiothreitol (DTT) and iodoacetic acid (IAA) were purchased from Sigma Chemicals (St. Louis, MO, USA) and stored at −20 °C. Other chemicals used were ethylenediaminetetraacetic acid (EDTA, Sigma), guanidine

Results

Previous efforts in developing a reversed-phase method for antibody characterization demonstrated good recovery, excellent peak shape, and minimal column interactions for most IgG1 antibodies of relatively low hydrophobicity [8]. However, significant tailing and reduced separation capabilities were encountered for more hydrophobic antibodies, including all IgG2 antibodies and certain IgG1 antibodies having particularly hydrophobic complementarity determining regions (CDR) as defined by the

Discussion

This optimized LC/MS method is a significant advancement for quality control and characterization of degradation products of therapeutics antibodies during their production and formulation. The method is able to resolve and quantify some forms of antibody heterogeneity and degradation products and in combination provide mass spectral data that can be used for characterization and identification. Although peptide mapping will continue to be a standard method for non-routine antibody analysis,

Acknowledgements

The authors thank Yu Zhang for stability samples, Himanshu Gadgil for data on GRAVY values, and David Brems and Susan Hershenson for their support and helpful discussions on the manuscript preparation.

References (46)

  • J. Bongers et al.

    J. Pharm. Biomed. Anal.

    (2000)
  • N.B. Afeyan et al.

    J. Chromatogr.

    (1990)
  • M.J. Rosok et al.

    J. Biol. Chem.

    (1996)
  • J.E. Battersby et al.

    J. Chromatogr. A

    (2001)
  • A. Beck et al.

    J. Chromatogr. B

    (2005)
  • T.M. Dillon et al.

    J. Chromatogr. A

    (2004)
  • J. Kyte et al.

    J. Mol. Biol.

    (1982)
  • K. Karch et al.

    J. Chromatogr.

    (1976)
  • B.E. Boyes et al.

    J. Chromatogr. A

    (1995)
  • Y. Chen et al.

    J. Chromatogr. A

    (2003)
  • C.T. Mant et al.

    J. Chromatogr. A

    (2003)
  • D.J. Kroon et al.

    J. Pharm. Biomed. Anal.

    (1995)
  • A.J. Cordoba et al.

    J. Chromatogr. B

    (2005)
  • O.H. Brekke et al.

    Immunol. Today

    (1995)
  • A. Wright et al.

    Trends Biotechnol.

    (1997)
  • R. Jefferis et al.

    Immunol. Lett.

    (1995)
  • C.T. Mant et al.

    J. Chromatogr. A

    (2003)
  • K.L. Richards et al.

    J. Chromatogr. A

    (1994)
  • K.L. Richards et al.

    J. Chromatogr. A

    (1994)
  • M.T. Hearn et al.

    J. Chromatogr.

    (1988)
  • K. Kato et al.

    J. Mol. Biol.

    (2000)
  • A.W. Vermeer et al.

    Biophys. J.

    (2000)
  • K. Benedek et al.

    J. Chromatogr.

    (1984)
  • Cited by (163)

    • Selective reversed-phase high-performance liquid chromatography method for the determination of intact SARS-CoV-2 spike protein

      2022, Journal of Chromatography A
      Citation Excerpt :

      Not only can antigens be separated by sequence but also post-translational modifications. As a result, RP-HPLC is a highly reproducible analytical tool widely used for a variety of biotherapeutics, such as purified protein, virus-like particles, and whole viruses [14–16]. However, as of this writing, only one brief application note with limited detail on S-protein analysis by RP-HPLC exists [17] and no published methods have characterized intact S-protein with its major subunits, S1 and S2.

    View all citing articles on Scopus
    View full text