An automated packed Protein G micro-pipette tip assay for rapid quantification of polyclonal antibodies in ovine serum
Introduction
The stringent cost and regulatory drivers now facing the pharmaceutical sector have prompted the investigation of a range of approaches to reduce development times and costs. For example, micro scale-down technologies have enjoyed a recent surge in interest and are becoming ever more widely established as viable techniques for accelerating process development [1], [2]. When used in conjunction with structured approaches such as factorial design, microscale studies offer the potential for the cost effective and thorough exploration of an experimental space, permitting far higher sample throughput than may be achieved in conventional bench-scale studies [3]. This allows novel products to move through pipelines more quickly while simultaneously enabling the type of thorough process description now encouraged by the Quality by Design initiative [4]. Several advances have been made in the microscale research field for operations such as fermentation [5], [6], microfiltration [7] and chromatography [8], [9]. These methods reduce the demands for feed material and increasingly are being implemented alongside automation technologies that enable many experiments to be conducted in parallel, thus allowing far more rapid, accurate and precise operation than that achieved manually. As a result, many data points can be gathered, permitting the generation of comprehensive response surfaces. Furthermore, since the quantities of resources consumed per microscale experiment are negligible (compared with more conventional laboratory- or pilot-scale studies), it becomes easier for a company to tolerate the development costs for studies that fail [10] or for therapeutic candidates that are abandoned following unsuccessful clinical trials.
For the full throughput benefits of these techniques to be realised, however, now requires significant reductions in analytical timescales [11]. Due to the intrinsically slow nature of some assay methods, these can become a major barrier to the high throughput screening of recovery and purification conditions [9] and depending upon the techniques involved, the time periods required to assess product or impurity concentrations can be far longer than the time needed to generate the samples in the first place. For example, in previous work, Chhatre et al. [3] found that the assay times needed to evaluate antibody concentrations in biological samples purified by automated microscale chromatography were an order of magnitude longer than the primary experiment itself. Predominantly this was due to the reliance upon HPLC, which requires extensive manual preparation, consumes significant quantities of buffers and involves long run times [9]. Alternatives to HPLC such as ELISA may require many replicates to achieve adequate confidence in the data and can involve long analysis times (e.g. due to overnight incubations), thus making it difficult to achieve precise and accurate results in a timely fashion. In the long term, the failure to address these concerns will result in analysis becoming a considerable obstacle to the implementation of high throughput process development.
Such issues call for novel high throughput assay techniques which need to be straightforward to use, necessitate little manual intervention and be commensurate with the small sample volumes available from the main scale-down experiment. Potential options available for rapid analysis include the Octet system (FortéBio, CA, USA), the Gyrolab (Gyros, Uppsala, Sweden) and the Bioanalyser (Agilent Technologies, CA, USA), all of which allow evaluation using small analyte volumes. These systems require separate instruments, thus entailing additional capital costs and lead times to train operators and develop robust assay protocols. Although automated liquid handlers can be expensive, if an experiment has been implemented on a robotic platform already, then execution of the assay on the same workstation can avoid further expenditure, thus improving its ‘return on investment’, and eliminating some of the complexity involved in connecting a robot to standalone analytical equipment. Although a period of familiarisation and training is needed for automation methods, once this understanding has been gained, the close integration of a microscale experiment and its analysis on a robot can reduce turnaround time between studies and so compress development timescales. To exemplify the analytical goals of such an approach, this paper describes a method which uses robotically controlled pipette tips packed with Protein G matrix as an alternative to antibody HPLC. To illustrate the principles involved, the technique was applied for the quantification of the antibody titre in ovine serum samples and the next section describes how the tip method was set-up, along with its comparison with a reference HPLC method.
Section snippets
Feed material
The crude hyperimmune ovine serum feedstock used in this study was obtained from BTG PLC (Blaenwaun, Ffostrasol, Llandysul, Wales, UK), which uses this material as the feed to manufacture FDA-approved bio-therapeutics such as CroFab™ [3]. A 3.6 L bag of serum at −20 °C was supplied by the company and thawed at ambient temperature for 8 h. The serum was agitated gently to ensure a homogenous composition, before being split into separate 150 mL lots for convenience of storage at −20 °C. Whenever
Determination of the number of Protein G load, wash and elution steps
The Protein G uptake curve for up to eight loading cycles with 1 mg/mL purified IgG is shown in Fig. 3. This indicates a steady reduction in the concentration until the fifth cycle, after which the values become more stable, indicating that the matrix is approaching an equilibrium concentration. Only a small reduction was observed between the fifth and eighth cycles and hence for reasons of throughput, it was decided that five loading cycles would be used for all further experimentation.
Conclusions
This paper has described the development of a high throughput alternative to Protein G HPLC for the quantification of polyclonal antibody in ovine serum. The method is based on the use of chromatography pipette tips packed with 40 μL of Protein G resin and automated on an eight-channel liquid handler. The tips were found to be reliable for detecting antibody concentrations in the 0.10–1.00 mg/mL range and the technique was used to calculate the immunoglobulin titre in crude hyperimmunised samples
Acknowledgements
The support of the Engineering and Physical Sciences Research Council Innovative Manufacturing Research Centre initiatives (IMRC) in Bioprocessing is acknowledged gratefully. The IMRC is part of The Advanced Centre for Biochemical Engineering, Department of Biochemical Engineering, University College London, with collaboration from a range of academic partners and biopharmaceutical companies (http://www.ucl.ac.uk/biochemeng/industry/imrc). The support of BTG PLC as well as Marc Wenger and
References (18)
- et al.
Curr. Opin. Biotechnol.
(2006) - et al.
J. Chromatogr. A
(2009) - et al.
Chem. Eng. Sci.
(2003) - et al.
J. Membr. Sci.
(2006) - et al.
J. Chromatogr. B
(2005) Trends Biotechnol.
(2002)- et al.
Process Biochem.
(2006) - et al.
J. Chromatogr. B
(2007) - et al.
Biotechnol. Bioeng.
(2009)