Continuous annular chromatography: General characterization and application for the isolation of recombinant protein drugs

• Published in: Biotechnology and bioengineering 2002-12, Vol.80 (5), p.559-568
• Main Authors: Vogel, Jens H., Nguyen, Huong, Pritschet, Markus, Van Wegen, Richard, Konstantinov, Konstantin
• Format: Article
• Language: English
• Subjects: Cells, Cultured; Chromatography - instrumentation; Chromatography - methods; continuous bioprocessing; Culture Media, Serum-Free; direct capture; factor VIII; Feasibility Studies; Industrial Microbiology - instrumentation; Industrial Microbiology - methods; ion exchange chromatography; peak wobbling; Pilot Projects; process integration; Quality Control; Recombinant Proteins - isolation & purification; Reproducibility of Results; Sensitivity and Specificity; Thromboplastin - genetics; Thromboplastin - isolation & purification
• ISSN: 0006-3592; 1097-0290
• DOI: 10.1002/bit.10411

Isolation of recombinant protein drugs from cell culture supernatant is usually performed in batch mode, even if the fermentation process itself is continuous. As a novel approach, continuous separation techniques like continuous annular chromatography (CAC) can be used for continuous isolation. The potential of CAC for industrial application is demonstrated by continuous isolation of rFVIII from cell culture supernatant in pilot scale (i.e., 144–288 l/day). Thirty‐fold concentration can be achieved at 94% yield, while purity is increased 3–5‐fold. For this a batch direct feed ion exchange chromatography method was adapted to a commercial preparative CAC system (P‐CAC). A headspace loading technique was used to maximize the concentration factor, while buffer incompatibility problems were addressed by a specifically modified inlet geometry. To allow sterile on‐line coupling to FVIII‐producing perfusion fermenters, an autoclavable pilot scale P‐CAC prototype was developed. General characterization of P‐CAC revealed a current limitation of the technology, i.e., variations in the outlet flow rates of up to ±20%. These flow variations are shown to be caused mainly by a nonuniform annular resin bed and in turn result in “peak wobbling,” i.e., the slight variation of peak position (up to ±4°) and shape (e.g., As = 0.9–1.4) as a specific function of column position. Some additional peak broadening, although less significant, is caused by a “peak oscillation” effect that results from the necessary segmentation of flow into discrete outlets. Both effects are only measurable if peaks are either monitored continuously or at least measured at multiple column positions. For isolation processes, these nonideal flow phenomena mean that more outlet streams have to be collected in order to achieve maximum yield and thus the achievable concentration factor is somewhat lower than the theoretical maximum. © 2002 Wiley Periodicals, Inc. Biotechnol Bioeng 80: 559–568, 2002.