ambr Automation Reduces Variance, Improves Data Quality

Presented by Mira Drohlshagen

Mira Drohshagen is an upstream process project engineer at Boehringer Ingelheim, Germany.

Several presentations at the Sartorius conference dissected various properties of the ambr® technology. Mira Drolshagen, a USP project engineer at Boehringer Ingelheim, evaluated the ambr250 platform. She stressed the applicability of the system for effective project planning with fewer setup time requirements. With its automated features, the ambr® technology can deliver low variance from sample to sample and high flexibility with very rapid time to completion. The critical question is whether it is an accurate model of large-scale bioreactor performance. To answer this query, Drolshagen and her colleagues performed a thorough investigation of ambr® performance.

In a series of studies with several different cell cultures protocols, they found that setup time was significantly reduced compared with nonautomated systems. Because the observed variation is quite low, a large number of replicates is not required.

Other features enhance the performance of the system. The ambr250 platform has capacity for 24 bioreactor vessels and can be managed by a single operator, whereas competitive systems can require extensive hands-on monitoring. It requires only one day for setup, and media fill can be executed two days before the initiation of an experimental run.

As a screening model, the highly flexible system provides a distinct advantage. For most projects the ambr250 platform shows good scalability. However, suitability as a scale-down model for process characterization studies needs to be established on a process-specific basis.

Drolshagen compared bubble size and distribution in mammalian and microbial cultures as a function of stirring speed. This is a critical problem because bubbling can cause protein denaturation and cell destruction due to surface tension. The denaturation process arises from the combined hydrophilic and hydrophobic nature of proteins causing unwinding and rearranging of the molecules. That rearrangement cleaves the protein’s structure and drives destruction of the molecules, which can lead to poor cellular growth performance. Indeed, denaturation is accompanied by precipitation of proteins out of solution and loss of functionality.

The Boehringer Ingelheim team observed that bubble size decreases with increasing stirring speed. Furthermore, for microbial vessels outfitted with a Rushton stirrer (a flat bladed turbine designed to maintain solids in suspension at low to medium viscosities), more small bubbles were detected than in the mammalian stirring vessel.

In conclusion, Drolshagen stated that the ambr system can be completely implemented as a screening tool, offering high flexibility and reproducibility. The high level of reproducibility translates to a much lower standard deviation for the data sets, which is a common property of many automated laboratory technologies that are rapidly replacing manual technician operations. She asserted that the many positive features of the ambr system — including the low variation from sample to sample, the reduction in setup time compared with other methodologies, and the substantial number of samples that can be run in parallel — make it ideal for optimizing parameters for bioprocessing scale-up.