April 15, 2013 (Vol. 33, No. 8)

Automating the Procedure via a Single-Use Mixing Platform

Low pH inactivation is commonly used in monoclonal antibody purification processes to inactivate large enveloped viruses. Adjusting the pH to achieve this as part of a single-use biomanufacturing workflow usually requires manual manipulation of the product and vessel by a skilled operator.

The adjustment comprises a number of steps including off-line sampling, pH measurement, and the accurate addition of the required amounts of acid and/or base—all the while ensuring that the pH does not exceed the required value.

In addition, manual interventions introduce additional risk to the process from operator error, microbial or viral contamination, pH overshoot, time-to-process deviations, and operator deviations.

Recording all the elements of these manual processes is an additional burden on the operator and a potential source of inaccuracy. Such inaccuracies can make identification of the cause of batch-to-batch variations more difficult, and a lack of electronic batch records, data storage, and trending in manual processing further complicates documentation and regulatory compliance.

Viral inactivation by low pH has been reliably demonstrated to inactivate more than >4log10 of large enveloped viruses (e.g., X-MuLV) in several commercial purification processes. This clearance step can be applied to monoclonal antibodies that have purification process steps that include a low pH step.

Relying on an automated single-use mixing platform that is able to contain, mix, and monitor the product, in addition to automating the pH adjustment steps, will simplify the viral inactivation steps and thereby mitigate the risks associated with manual adjustment.

To investigate viral inactivation using this method, two studies were performed to assess the performance of such a device, the Xcellerex™ XDUO™ 500 L mixer. The first study assessed the mixing capability of the system, vital for effective pH adjustment; the second investigated the ability to automate pH adjustment within the required parameters to successfully perform the viral inactivation step in a purification process for a monoclonal antibody.

Automated 500 L Mixing Study

A 500 L XDM™ Quad Plus bag was installed into the 500 L mixer and pH probes were fitted into the appropriate ports. Although the probes were installed, data from these probes was not collected for this experiment. To facilitate placement of pH probes, posts were fixed inside the mixing vessel through incisions made through the top of the bag.

One probe was inserted through the central port in the top of the bag, as shown in Figure 1. To sample the pH from each part of the mixing vessel, probes were positioned in the corners as well as in the center of the vessel. pH measurements were taken using data-logging software and pH readings were collected at one second intervals from each of the probes.

Figure 1. For the XDUO 500 L mixing experiment a total of 8 pH probes were placed throughout the vessel as shown. pH measurements were taken at one second intervals. Agitator speed was varied at intervals of 50 from 50 rpm through 200 rpm. Clockwise (down) and counterclockwise (up) directions were tested at each agitator (rpm) interval.

The 500 L mixer was filled to 450 L with 20 mM glycine, pH 5.5. Time between each pH spike and return to uniformity was determined for each mixing speed for both clockwise and counterclockwise directions. pH for each probe was traced following a spike with acid or base into the vessel. Volume of peak solution was 10 mL of 10 N sodium hydroxide (base), or 30 mL of 6 N hydrochloric acid.

pH peak solution was measured volumetrically with a 50 mL conical tube. For this test, a sample of acid or base was poured in through the top port on the bag. pH readings from each probe were collected until a stable baseline returned. Variables were agitator speed (rpm) and direction of rotation, clockwise (downflow) vs. counterclockwise (upflow).

The pH was tracked for 120–160 seconds following each addition to measure re-equilibration time. pH data were saved and imported into Microsoft Excel. Probes used were not calibrated prior to this experiment as the pH data were used only to determine the time interval for return to stable baseline following each pH peak (Table 1).

Table 1: Average time to equilibration

Automated pH Adjustment Study

Typically, manual product pH adjustment requires operators to withdraw multiple samples from a product bag or vessel for offline pH measurement and pH adjustment. The required volume of pH adjustment solution to reach the desired pH of 3.0 is calculated offline before 80% of the volume is added to the product pool, either manually with a syringe through a sample port, or for larger volumes, with a peristaltic pump.

Consequently, the pH adjustment solution is added rapidly and without constant mixing, rather than gradually with constant mixing. This may not be ideal for all proteins, as mixing is not controlled during the addition. Following buffer addition, the product is mixed for 5–10 minutes. Next, samples are taken for offline pH measurement. Based on test results, the remaining acid or base is either added in full or in part. Once again, the product is mixed manually and more samples are taken to measure pH until target pH is reached.

Once the set-point pH is reached, the viral inactivation hold time begins. Typically this is a 60 min hold followed by pH adjustment up to 5.0–7.0 following the same method, and incurring the same process risk, as previously mentioned. Reducing VI step duration is prudent as it limits/minimizes product exposure to a low pH environment as some proteins contain acid labile groups, and even relatively mild acid treatment may cause irreversible loss of function.

Automated pH Adjustment

A 500 L XDM Quad Plus bag was installed into the 500 L mixer and pH probes were fitted into the appropriate ports. Test volume chosen was 125 L. 128 kg of water (RODI) was added to the mixer using a Levitronix BPG-600 pump. 1.30 L 2 M glycine (pH 4.4 stock) was measured volumetrically and then added through the top port of the Quad Plus bag. For each run, the mixer’s agitator speed was set to 100 rpm in the clockwise (CW down) direction.

Once the pH adjustment solutions were set up next to the mixer, 1/8” ID C-flex tubing was installed on both addition pumps: pump 01 (200 mM HCl), pump 02 (500 mM Tris-base). A set-point pH of 3.0 was entered into the system and when this was achieved, the agitation was held at 100 rpm for 5 min to demonstrate stable pH was met. Next the set-point was changed to pH 7.0 and once this was achieved, vessel agitation was again held at 100 rpm for 5 min to demonstrate stable pH was achieved.

A sample was taken from the mixer for offline pH measurement using a calibrated pH unit to check the accuracy of the mixer’s pH probes.

Data were captured and imported into Microsoft Excel using the system’s Data Historian software as provided. Table 2 summarizes the process parameters as recorded.

Starting solution used was 20 mM glycine, pH 4.25 in a starting volume of 125 L (Figure 2). Automated pH adjustment was performed with the mixer and the following inputs were entered into the unit:

1st pH set point = 3.0, deadband = 0.05; 2nd pH set point = 7.0, deadband = 0.05. For pH 3.0 adjustment, 2.69 L of 200 mM HCl was added over 31 mins (final pH was 3.0). For the pH 7.0 set point, 1.70 L of 500 mM sodium hydroxide was added over 19.5 minutes (final pH was 6.98).

Table 2: Summary of automated pH adjustment for 125 L performed using the XDUO mixer.


In these studies, the XDUO (GE Healthcare) single-use mixing system has demonstrated how accurate, automated pH adjustment, within accepted industry process standards, can be achieved with full capture of the process data to a historian platform.

Using a standalone, easy-to-use automated mixing platform for viral inactivation reduces or eliminates the risks associated with manual execution and simplifies the process. Such an approach has numerous advantages in pH adjustment operations for biomanufacturing:

  • Reduced risk of process deviations from operator error (automation)
  • Increased time savings by eliminating deviations and manual batch records (automation)
  • Reduced risk of contamination and protection of the product and the operator (single-use bags, closed systems)
  • Increased plant efficiency with reduction of equipment footprint requirements (flexible single-use equipment).

Figure 2: XDUO mixer automated pH adjustment of 125 L of 20 mM glycine, pH 4.25

Joseph Makowiecki ([email protected]) is manager process development and Heather Mallory is senior research associate process development, downstream at GE Healthcare Life Sciences.

Previous articleOn Supreme Court Docket: Whose Gene Is It Anyway?
Next articleTransplanted Bioengineered Kidneys Successfully Produce Urine