Extreme Scale-Down Approaches for Rapid Chromatography Column Design and Scale-Up During Bioprocess Development - Measurement, Monitoring, Modelling, and Control of Bioprocesses - page 129

reducing analyte volumes, microfluidics may help to facilitate integrated

micro-approaches for sequences of process steps;
• Automating assays wherever possible can improve data consistency and release
operator time to perform other tasks such as data evaluation. Robots can also be
used to perform preparative steps such as buffer make-up, potentially on the
same system used to execute the main microscale experiment.

One example where assay automation and parallelisation can be beneficial lies
in antibody titre determination by Protein G HPLC. HPLC systems work in a serial
fashion, meaning that the timescales involved can become long. Although com-
mercial equipment is available for processing many samples at the same time, this
requires standalone and costly equipment. Especially if a high-throughput screen is
conducted on a robotic system, then the use of a more integrated form of analysis
would be preferential such that the analysis takes place on the same robot
(reducing down-time by starting analysis directly after the main experiment and
thus returning data to the end user more rapidly). A recent study developed a high-
throughput antibody assay as a replacement for HPLC using an eight-channel
liquid handling robot to manipulate chromatography pipette tips containing 40 lL
of Protein G [ 5 ]. The study evaluated a number of analytical properties, including
range and detection limits, linearity, accuracy, precision and specificity. After this,
the method was tested by quantifying the titre in an ovine feedstock used com-
mercially for making an approved therapeutic product. The average titre calculated
using the chromatography tips was comparable to that determined by HPLC, but
the eight-channel robotic tip approach delivered results in less than 40 % of the
HPLC time. The potential for further time savings by using higher-capacity robots
was also identified, with a 96-channel pipetting system offering the possibility of
saving approximately 90 % of the time. Integrating the analytical set-up together
with the main scale-down experiment such as robotic systems thus offers the
potential for 'walkaway' automation.

5.2 Experimental Design Methods

5.2.1 Factorial and Iterative Designs

Appropriate experimental design techniques are needed for HTS to maximise
search space coverage, although this must be balanced with the amount of labo-
ratory effort required. Although it is possible to carry out 'brute force' experi-
mentation by HTS, this would consume large quantities of materials, time and
money and result in a considerable analytical workload. A better approach is to
reduce the experimental requirements to the minimum needed to deliver the
required process knowledge in a timely fashion. Thus test conditions can be
chosen to provide approximately the same level of information as a brute force
design but with a far lower experimental burden; For example, factorial designs