Downstream processing of human antibodies integrating an extraction capture step and cation exchange chromatography

https://doi.org/10.1016/j.jchromb.2008.11.014Get rights and content

Abstract

In this paper we explore an alternative process for the purification of human antibodies from a Chinese hamster ovary (CHO) cell supernatant comprising a ligand-enhanced extraction capture step and cation exchange chromatography (CEX). The extraction of human antibodies was performed in an aqueous two-phase system (ATPS) composed of dextran and polyethylene glycol (PEG), in which the terminal hydroxyl groups of the PEG molecule were modified with an amino acid mimetic ligand in order to enhance the partition of the antibodies to the PEG-rich phase. This capture step was optimized using a design of experiments and a central composite design allowed the determination of the conditions that favor the partition of the antibodies to the phase containing the PEG diglutaric acid (PEG-GA) polymer, in terms of system composition. Accordingly, higher recovery yields were obtained for higher concentrations of PEG-GA and lower concentrations of dextran. The highest yield experimentally obtained was observed for an ATPS composed of 5.17% (w/w) dextran and 8% (w/w) PEG-GA. Higher purities were however predicted for higher concentrations of both polymers. A compromise between yield and purity was achieved using 5% dextran and 10% PEG-GA, which allowed the recovery of 82% of the antibodies with a protein purity of 96% and a total purity of 63%, determined by size-exclusion chromatography. ATPS top phases were further purified by cation exchange chromatography and it was observed that the most adequate cation exchange ligand was carboxymethyl, as the sulfopropyl ligand induced the formation of multi-aggregates or denatured forms. This column allowed the elution of 89% of the antibodies present in the top phase, with a protein purity of 100% and a total purity of 91%. The overall process containing a ligand-enhanced extraction step and a cation exchange chromatography step had an overall yield of 73%.

Introduction

Biopharmaceutical products, including proteins, nucleic acids and other bioproducts, have application in several focus areas, including vaccination, immunization, oncology, autoimmune, cardiovascular, inflammatory and neurological diseases. The market of biopharmaceutical products is one of the fastest growing segments of the pharmaceutical industry, and as the efficacy of traditional drugs is decreasing and population is aging, the demand for new therapies, that will allow the treatment of the 21st century diseases, is increasing. Industry is thus focusing on developing new biotechnology-derived products. The annual sales of several biopharmaceuticals have surpassed the billion dollar mark [1]. In 2006, the global market for biopharmaceuticals was estimated to be worth around $50 billion [2] and is expected to overpass the $90 billion by 2010 [3]. With the completion of the primary DNA mapping of the human genome and the progress made in high-throughput technology for drug discovery, a remarkable growth in the development of therapeutic products was observed, with more than 500 products in active clinical trials [2].

Nevertheless, the demand for large quantities of therapeutic proteins for the treatment of diseases requiring high doses and/or chronic administration has brought several concerns about a potential shortage of the manufacturing capacity, which, in turn, intensified the pressure to improve cell culture productivity [4]. Advances in molecular biology have allowed the increase of cell line productivity, by optimizing media formulations and feed delivery strategies. Process control, bioreactor design, and host cell engineering have also been significant factors in improving upstream processes [4]. Antibody titers, for example, have increased from few milligrams a decade ago to multigrams per liter today, with reports of titers reaching 15 g/l [5]. Such increase has been attributed to an increase in both cell density and duration of cell viability [6].

Nevertheless, while upstream productivity is increasing in line with demand, improvements in downstream processes have been neglected, as biopharmaceutical companies are reluctant to change long established processes for new ones. This is resulting in a production bottleneck that is shifting the costs of production downstream [7]. In fact, downstream processes of biopharmaceuticals can account for up to 80% of total manufacturing cost [8] and has been a long-standing market barrier. As a result, the need to improve process economics and efficiency, to reduce costs and to meet the increasingly demand of quality for market approval is forcing the development of more efficient and cost-effective separation and purification methods in order to keep up with the upstream gains.

The platform approach for the downstream processing of antibodies usually encompasses three chromatographic steps, a Protein A affinity capture step and two chromatographic polishing steps to remove host cell proteins, high molecular weight aggregates, low molecular weight clipped species, DNA and leached Protein A that remain after the capture step [9]. Protein A chromatography has become the traditional choice for antibody capture not only due to its specificity and purity levels obtained but also due to its ability to handle unconditioned feeds straight from the bioreactor. Nevertheless, Protein A chromatography has several inherent limitations because of its high cost and leachability.

Non-chromatographic methods could be a valuable alternative to some of the chromatographic steps as long as the final purification is not compromised. Several promising alternatives have been described in the literature, including, affinity precipitation [10], [11], liquid–liquid extraction [12], [13], high performance tangential flow filtration [14], membrane chromatography [15], [16], high gradient magnetic fishing [17] and crystallization [18]. These alternative technologies, and in particular, aqueous two-phase partitioning, aim at high throughput and seek to avoid problems associated with cost, capacity and diffusional limitations encountered with most chromatographic supports [19].

Aqueous two-phase systems (ATPSs) composed of dextran and polyethylene glycol (PEG) have been shown suitable for the capture of human antibodies from a CHO cell supernatant, as long as the affinity of the antibody for the PEG-rich phase is enhanced by the addition of a ligand [20], [21], [22]. The modification of polyethylene glycol terminal hydroxyl groups with an amino acid mimetic ligand (glutaric acid) enhanced the partitioning of antibodies to the PEG-rich phase from 23% in the unfunctionalized system to more than 90% in the glutaric acid functionalized system [22]. In this paper, we have used a design of experiments to optimize the ATPS extraction step, in terms of system composition, by response surface analysis. This approach allows studying the effect of polyethylene glycol diglutaric acid (PEG-GA) and dextran concentration and their possible interaction without varying one condition at the time, and consequently reducing the number of required experiments. We also explore the possibility of integrating the ATPS extraction step with a cation exchange chromatography (CEX) step without any pre-conditioning step. This would not only simplify the process but would also reduce the overall costs, especially at large scale.

Section snippets

Materials

Human IgG for therapeutic administration (product name: Gammanorm) was purchased from Octapharma (Lachen, Switzerland), as a 165 mg/ml solution containing 95% of IgG. The cell supernatant was produced and delivered by ExcellGene (Monthey, Switzerland) and contains a human IgG1, directed against a human surface antigen. An ExcellGene proprietary serum-free medium was used for production. Phenol red has been added to the medium as a pH indicator. According to Excellgene the antibody titer was 110 

Aqueous two-phase extraction optimization

The selective capture of human antibodies from a CHO cell culture supernatant by aqueous two-phase extraction using a polymer functionalized with an amino acid mimetic ligand (polyethylene glycol glutaric acid, PEG-GA) was optimized by response surface methodology (RSM), which comprises three general steps: the experimental design; modeling and statistical validation; and optimization [24], [25], [26].

Conclusion

We have developed a process for the purification of human antibodies that allows the successful integration of affinity aqueous two-phase extraction with cation exchange chromatography. This process has an overall yield of 75% and allows the recovery of the antibodies with a total purity of 91% and about 100% protein purity. The capture of IgG from the CHO cell supernatant was successfully accomplished with a PEG/dextran system in which PEG hydroxyl groups were modified with an amino acid

Acknowledgements

This work was carried out within the European project AIMs (contract No. NMP3-CT-2004-500160), supported by funding under the Sixth Research Framework Program of the European Union. A.M.A., P.A.J.R. and I.F.F acknowledge “Fundação para a Ciência e Tecnologia” for the fellowships BPD/41274/2007, BD/25040/2005 and BD/38941/2007, respectively.

References (27)

  • A.A. Shukla et al.

    J. Chromatogr. B

    (2007)
  • A. Kumar et al.

    Prog. Polym. Sci.

    (2007)
  • A.C.A. Roque et al.

    J. Chromatogr. A

    (2007)
  • M. Rito-Palomares

    J. Chromatogr. B

    (2004)
  • R. van Reis et al.

    Curr. Opin. Biotechnol.

    (2001)
  • D. Josic et al.

    J. Chromatogr. B

    (2001)
  • C. Boi

    J. Chromatogr. B

    (2007)
  • P.A.J. Rosa et al.

    J. Chromatogr. A

    (2007)
  • A.M. Azevedo et al.

    J. Biotechnol.

    (2007)
  • D.K. Follman et al.

    J. Chromatogr. A

    (2004)
  • R.J.Y. Ho et al.

    Biotechnology and Biopharmaceuticals-Transforming Proteins and Genes into Drugs

    (2003)
  • Frost & Sullivan

    Biotechnology Up-stream and Down-stream Processing Technology Analysis (Technical Insights)

    (2006)
  • Visiongain

    Therapeutic Proteins

    (2005)
  • Cited by (0)

    View full text