Introduction
Viral safety and viral clearance procedures are mandatory requirements for protein therapeutics (ICH, 1998). Low-pH viral inactivation (VI), operated in either batch or continuous mode, often follows the Protein A capture step in downstream purification due to the acidic condition of Protein A eluate (PAE) (Parker et al., 2018). A typical low-pH VI operation includes an acidification (VIA) step, where PAE is adjusted to around pH 3.6 and held for up to several hours to achieve sufficient viral inactivation (Brorson et al., 2003; Mattila et al., 2016); and a neutralization (VIN) step, where the VIA pool is adjusted to a pH value close to neutral for further downstream processing (Shukla, Hubbard, Tressel, Guhan, & Low, 2007).
Even though most monoclonal antibodies (mAbs) are stable under low-pH conditions during VIA, it was reported that low-pH conditions induced denaturation and subsequent aggregation of proteins, which are pH sensitive phenomena (Brorson et al., 2003). Protein aggregates have been indicated as one of the main causes for loss of therapeutic activity and adverse immune reactions (Rosenberg, 2006). Therefore, the presence of protein aggregates in biopharmaceutical products has become a main concern for the biopharmaceutical industry and regulatory agencies (Filipe, Kükrer, Hawe, & Jiskoot, 2012). Preventing and monitoring the formation of product aggregates is critical during process development and manufacturing of mAb products (den Engelsman et al., 2011).
A number of studies have shown that incubation at low pH induces the formation of an aggregate precursor with molten globule-like structures (Bychkova, Berni, Rossi, Kutyshenko, & Ptitsyn, 1992; Muzammil, Kumar, & Tayyab, 1999; Redfield, Smith, & Dobson, 1994). Filipe et al. (2012) described aggregation of an IgG1 under low-pH stress, where the protein molecules converted into transient, partially unfolded monomers when pH was changed from 6.0 to 1.0. The modified monomers then either refolded back to the native state or initiated an aggregation process after changing pH back to 6.0 (Filipe et al., 2012). Skamris et al. (2016) characterized the oligomerization kinetics at pH 3.3 and the reversibility upon neutralization for three immunoglobulin-G proteins (IgGs) representing typical IgG1, IgG2 and IgG4, respectively. Results revealed distinct solution behaviors among the three IgGs. At acidic pH, IgG1 retained monomeric, whereas IgG2 and IgG4 exhibited the formation of transient and partially unfolded monomers, namely oligomers. Subsequent neutralization of PAE caused IgG2 oligomers to partially reverse to the monomeric state, whereas, IgG4 oligomers tended to aggregate (Skamris et al., 2016).
The pH of PAE pool is typically adjusted to 3.5 to 3.8 by addition of an acid solution (Brorson et al., 2003; Shukla et al., 2007) and the VIA condition expectedly should not cause mAb aggregation under the sufficient mixing condition. It held true in bench scale experiments. Using IgG4-N1 PAE, the VI process was developed at bench scale in glass beakers or in Eppendorf tubes up to 50 mL, where sufficient mixing was provided to create a homogeneous pH environment (Jin et al., 2019). Unlike setting for the bench scale VI operation, the flow velocity in a pilot/production scale mixer is usually limited to the laminar regime to minimize shear forces on the product (Parker et al., 2018). This condition is similar to the well characterized mammalian cell culture in large scale bioreactors, where insufficient mixing could lead to pH heterogeneities, characterized as low- or high-pH zone (lower or higher than pH set point) near acid or base addition points (Bylund, Collet, Enfors, & Larsson, 1998; Langheinrich & Nienow, 1999 ; Lara, Galindo, Ramírez, & Palomares, 2006; Xing, Kenty, Li, & Lee, 2009). This led to a hypothesis that a localized low-pH zone (≤ pH 3.3) might exist during VIA caused by poor mixing condition.
Computational fluid dynamics (CFD) modeling has become a useful tool to evaluate the pH heterogeneities in the VI operation to avoid at-scale tests using costly PAE. CFD modeling was employed to quantify residence time distribution in a continuous VI tubular reactor. Based on the criterion of minimum residence time for the tracer fluid elements in the reactor, the geometric and operating parameters, such as pipe length, flow direction, and the use of secondary flows, were optimized to enhance radial mixing (Parker et al., 2018). CFD modeling was also employed to simulate the steady state flow field velocity profiles of the pulse tracer experiments, which helped researchers to understand the complex flow behavior of the reactor for low-pH VI (Francis & Haynes, 2009; Kelly, 2008; Lode, Rosenfeld, Yuan, Root, & Lightfoot, 1998; Manninen, Gorshkova, Immonen, & Ni, 2013). However, those approaches were not able to predict quantitatively HMW level after the VI operation.
Most recently, Jin et al. (2019) employed CFD modeling to demonstrate the presence of low-pH (≤ pH 3.3) zone near the acid addition point of the mixer, which caused formation of a partially unfolded IgG4 monomer during VIA and led to product aggregation during VIN (Jin et al., 2019). This work focus on the details of the modeling methodology. Furthermore, based on the CFD modeling outcomes and constants obtained from protein titration experiments, a HMW prediction model was developed to quantify HMW level after the VI operation in the confined geometry of mixing vessels, where factors such as agitation speed, acid addition rate, and protein concentration were considered. The model was used for troubleshooting an initial scale-up run failed with excessive HMW formation and guiding the subsequent 12 pilot scale (50-200L) runs. This work aimed to explore the application of the HMW prediction model to facilitate the scale up of the low-pH VI process directly from bench to pilot/production scale.