Adapted Ultra Scale‐Down Approach for Predicting the Centrifugal Separation Behavior of High Cell Density Cultures

Abstract

The work presented here describes an ultra scale-down (USD) methodology for predicting centrifugal clarification performance in the case of high cell density fermentation broths. Existing USD approaches generated for dilute systems led to a 5- to 10-fold overprediction of clarification performance when applied to such high cell density feeds. This is due to increased interparticle forces, leading to effects such as aggregation, flocculation, or even blanket sedimentation, occurring in the low shear environment of a laboratory centrifuge, which will not be apparent in the settling region of a continuous-flow industrial centrifuge. A USD methodology was created based upon the dilution of high solids feed material to approximately 2% wet wt/vol prior to the application of the clarification test. At this level of dilution cell-cell interactions are minimal. The dilution alters the level of hindered settling in the feed suspensions, and so mathematical corrections are applied to the resultant clarification curves to mimic the original feed accurately. The methodology was successfully verified: corrected USD curves accurately predicted pilot-scale clarification performance of high cell density broths of Saccharomyces cerevisiae and Escherichia coli cells. The USD method allows for the rapid prediction of large-scale clarification of high solids density material using millilitre quantities of feed. The advantages of this method to the biochemical engineer, such as the enabling of rapid process design and scale-up, are discussed.

Assessment of the manufacturability of Escherichia coli high cell density fermentations

“…At this level of dilution, cell-cell interactions leading to hindered settling are minimal. 11 The parameters volume per well (0.5-1.5 mL) and centrifugation times (5 and 10 min) were varied to achieve V/ctR (¼ Q/cR) values ranging from 7.6 Â 10 À9 to 3.9 Â 10 À8 m s À1 . After centrifugation, the supernatant was aspirated ensuring the pellet was not disturbed (volume per well in milliliter [volume removed in milliliter]): 0.5 (0.3); 1.0 (0.75); 1.5 (1.0), for analysis by slowly pipetting, while moving the pipette downward along the well wall.…”

“…[11][12][13] The difficulty of the fermentation, with relation to its high oxygen demands 14 and process complexity, due to its multiphase nature (i.e., batch phase followed by fed-batch phase), makes this high cell density cultivation 15 a good candidate to assess the use of scale-down techniques in conjunction with analytical techniques such as particle size distribution (PSD) and viscometry to predict the effects of harvest time on downstream processing performance.…”

“…11 This has been extended to microscale using microwell plates allowing for automation and high throughput. 20 Wenger et al 21 described the use of adaptive focus acoustics (AFA), a technique that delivers highly focused acoustic energy through the use of ultrasonication, as a microscale yeast cell disruption mimic.…”

Influence of pH control in the formation of inclusion bodies during production of recombinant sphingomyelinase-D in Escherichia coli

“…IBs are dynamic reservoirs that contain a large amount of recombinant protein, various host proteins like chaperones, among other components of the cytoplasm [4,8]. IBs are highly hydrated dense particles of porous structure [9,10], their surface varies from rough to smooth [6], and their size is normally in the range of 50 to 700 nm, having spherical, cylindrical or ellipsoidal teardrop shapes [10][11][12][13][14][15].…”

Intelligent Automation Platform for Bioprocess Development

“…Initially developed for drug discovery research where thousands of compounds are tested to identify drug candidates, the concept of high throughput approach has been adapted in the last decade to reduce the time and cost in bioprocess development. A series of unit operation ultra scale down (USD) technologies have been developed in which a microscale device is used to mimic the key features of a large scale operation (Tustian et al 2007;Neal et al 2003;Ma et al 2009 Therefore when the dataset is large, researchers often take substantial time for data analysis, process evaluation and process understanding. The new experiment design requires the user's inputs which come from the process understanding, e.g.…”