An ultra scale‐down approach to study the interaction of fermentation, homogenization, and centrifugation for antibody fragment recovery from rec E. coli

Abstract

Escherichia coli is frequently used as a microbial host to express recombinant proteins but it lacks the ability to secrete proteins into medium. One option for protein release is to use high-pressure homogenization followed by a centrifugation step to remove cell debris. While this does not give selective release of proteins in the periplasmic space, it does provide a robust process. An ultra scale-down (USD) approach based on focused acoustics is described to study rec E. coli cell disruption by high-pressure homogenization for recovery of an antibody fragment (Fab') and the impact of fermentation harvest time. This approach is followed by microwell-based USD centrifugation to study the removal of the resultant cell debris. Successful verification of this USD approach is achieved using pilot scale high-pressure homogenization and pilot scale, continuous flow, disc stack centrifugation comparing performance parameters such as the fraction of Fab' release, cell debris size distribution and the carryover of cell debris fine particles in the supernatant. The integration of fermentation and primary recovery stages is examined using USD monitoring of different phases of cell growth. Increasing susceptibility of the cells to disruption is observed with time following induction. For a given recovery process this results in a higher fraction of product release and a greater proportion of fine cell debris particles that are difficult to remove by centrifugation. Such observations are confirmed at pilot scale.

Periplasmic expression in and release of Fab fragments from Escherichia coli using stress minimization

“…We have demonstrated that stress minimization by decreasing culture temperature and inducer concentration can be used to improve not only recombinant protein yields and folding but also correct subcellular targeting and overall cell integrity for subsequent cell harvest and protein release steps . Careful balancing of increases in inducer concentration and decreases in growth and induction temperature allowed optimization of Fab D1.3 yield.…”

“…In order for periplasmic release to succeed, bacterial viability and physical integrity must be maintained upon harvest. Metabolic stress in bacteria generating recombinant proteins such as periplasmically‐targeted Fab fragments can cause cell lysis during fermentation or fragility leading to lysis during cell harvesting in industrial centrifuges or subsequent processing steps . This eliminates the advantages of periplasmic expression as cytoplasmic proteins, membrane components and DNA are all released, complicating subsequent purification of the target recombinant protein.…”

Microscale disruption of microorganisms for parallelized process development

“…Baumann et al and Ladd Effio et al used a parallelized system for eight samples with volumes between 1 and 5 mL for quantitative comparison of high‐throughput upstream screenings. Wenger et al and Li et al describe adaptive focused acoustics (AFA) as an advanced sonication method. Higher frequencies lead to shorter acoustic cycles resulting in reduced time for bubble growth and therefore smaller bubbles with improved mixing .…”

“…The extent of cell disruption and the method itself affect the composition and biophysical and biochemical properties of the homogenate such as viscosity, density, particle size distribution, host cell protein, host cell DNA and endotoxin level. These properties in turn impact the subsequent purification steps such as centrifugation , filtration and chromatography and finally the result of the whole purification process chain .…”

“…host cell protein, endotoxin, DNA), product titer, and crucial process parameters such as viscosity and size of cell debris. Only considering these parameters will enable successful transfer to industrial scale at speed . A method suited for integration into a process development platform must further be suitable for parallelized experimentation.…”