Application Note

Rapid, reliable quantitation of Fc-Fusion protein in cell culture supernatants

Download PDF    

Introduction

Kanvasri Jonnalgadda, Lam R. Markely, Scott Estes, Shashi Prajapati, High-Throughput Analytical Group (HTAG)-Cell Culture
Development, Biogen IDEC, Cambridge, MA
Rashi Takkar, Sriram Kumaraswamy, Applications Development, ForteBio, Fremont, CA

Background

The cell line development group at Biogen IDEC needed a robust assay for the measurement of Fc-fusion protein (Protein 1) in crude cell culture supernatants. The group had historically used HPLC for protein quantitation during screening and selection of promising mammalian clones at every scale-up step, from 96- well microplates to 3L-bioreactors. They wanted to replace the HPLC method due to its low throughput, cumbersome sample processing, and long run times with a higher-throughput alternative that provided accurate, reliable results.

Octet platform solution

The Octet® QK384 instrument was evaluated as an alternative to HPLC due to its many advantageous features and suitability for screening proteins in cell culture supernatant fluids. Octet systems have the ability to analyze crude samples, allowing users to bypass sample pre-processing. In addition, 96 samples could be analyzed in less than 30 minutes, expediting screening that took more than 19 hours to complete by HPLC. A higher throughput, automated screening workflow with significantly reduced analyst involvement was achieved via integration of the Octet system and a PerkinElmer (formerly Caliper Life Sciences) Sciclone robot. This document summarizes the performance of Biogen IDEC’s Fc-fusion protein assay.

Assay development and results

METHOD

An Octet QK384 instrument integrated with a PerkinElmer Sciclone robot was used for all experiments. Samples were analyzed in 96-well, black, polypropylene, flat-bottom microplates from Greiner (Part No. 655209). Protein A and Protein G biosensors were selected because of their high specificity and selectivity for Fc-fusion protein detection in supernatants. Default settings of a 2-minute read time and 400 RPM shake speed were selected.

OCTET ASSAY PROVEN ROBUST AND RELIABLE

During assay development, standard curves were constructed by diluting the purified bulk drug substance at concentrations from 0 to 100 μg/mL in cell culture media (CCM). Assays were performed in triplicate, using identical well locations in three independent experiments. The average coefficient of variation (CV) for the entire standard curve range was less than 9% for the three combined runs (Figures 1A and 1B). A dynamic range of 1-100 µg/mL was found to be optimal, with acceptable recovery values and CVs. Standard curves were also assayed in different well locations in three independent experiments to account for any well-to-well variability (Figures 2A and 2B). Assay performance was similar across all well locations, demonstrating good consistency across the plate.

Standard curves of the Fc-fusion

Standard curves of the Fc-fusion protein two

Standard curves of the Fc-fusion protein three

Figure 1A: Standard curves of the Fc-fusion protein assay using Protein G biosensors, obtained in three independent experiments using the same 96-well microplate well locations (BR = Binding Rate). The calculated recovery values are shown in the plate map.

Pooled standard curve data generated at same well locations

Figure 1B: Pooled standard curve data generated at same well locations in three independent experiments. The results were highly consistent across the entire plate, with CVs for all three combined runs below 10%.

Standard curves were generated in different well locations

Standard curves were generated in different well locations two

Standard curves were generated in different well locations three

Figure 2A: Standard curves were generated in different well locations in three independent experiments to check consistency of results.

Pooled standard curve data generated at different well locations

Figure 2B: Pooled standard curve data generated at different well locations in three independent experiments. The results were highly consistent across the entire plate, with a CV for all three combined runs below 10%.

Cell Culture Media Had No Effect On Results

Interference by matrix components was checked to ensure the assay could be performed on all upstream process samples with minimal pre-processing. Standard samples were assayed in various dilutions of cell culture media: 8-fold (12.5% CCM), 4-fold (25% CCM), 2-fold dilutions (50% CCM), and neat (100% CCM) (Figure 3). The four standard curves were then overlaid and overall variability was determined at each concentration. The assay was found to tolerate up to 100% CCM, with comparable recoveries of spiked standards in all dilutions and an overall CV of <8%.

The assay tolerated neat

Figure 3: The assay tolerated neat (100%) CCM. 100%, 50%, 25% and 12.5% CCM had comparable recoveries of the standard curve with overall combined CV <8%.

RESULTS WERE ACCURATE

The accuracy of the Octet assay was evaluated by spiking Fc-fusion protein into CCM and host cell protein (HCP) matrices at known concentrations in the low (10 μg/mL), mid (40 μg/mL) and high (80 μg/mL) range of the standard curve (Figure 4). Spike-in recoveries for the high, mid and low controls in neat CCM were 95%, 104%, and 109% respectively. Recoveries of all control samples in 2 (50%), 4 (25%) and 8 (12.5%) fold dilutions of the HCP matrix were close to 100%, demonstrating excellent accuracy. Recoveries of the control samples in neat HCP matrix were around 122%, showing values higher than acceptable levels. In all subsequent assays, standard and unknown samples were run with at least a 2-fold dilution. Minimization of matrix interference and consistency between matrices used in standard and unknown samples are important for getting accurate results.

Fc-fusion protein was spiked at three different concentrations

Figure 4: Fc-fusion protein was spiked at three different concentrations in CCM and HCP matrix to determine the accuracy of the assay. 100% recovery was achieved in neat CCM and diluted HCP matrices, however, 122% recovery was seen in undiluted HCP.

Good Dilution Linearity

Dilution linearity is an important feature of any assay used to screen samples containing a wide range of protein concentrations. Fc-fusion protein samples at low (S1), high (S2) and medium (S3) range concentrations were diluted 5, 10, 20, 80 and 160-fold and analyzed in triplicate (Figure 5). The CV for back-calculated values of the low concentration sample S1 was <7%. Corresponding CV values for high concentration sample S2 and medium concentration sample S3 were <5%. These results indicate excellent dilution linearity and recoveries for the Fc-fusion protein in cell culture supernatant.

Dilution linearity was confirmed by diluting three concentrations

Figure 5: Dilution linearity was confirmed by diluting three concentrations of the Fc-fusion protein to various levels. Assays were performed in triplicate.

Good Correlation Between Octet Data And Hplc

When evaluating the performance of Octet assays to complement or replace HPLC protocols, users generally perform comparative studies. If there are significant differences in sample preparation and handling between the two methods, data correlation may be affected. For this experiment, all samples to be analyzed on both platforms were centrifuged and filtered prior to analysis. In the Octet assay, early cell culture samples were run with at least a 2-fold dilution in CCM, and bioreactor samples were diluted 20- to 40-fold in CCM to ensure final analyte concentrations were within the assay’s dynamic range (below 100 μg/mL). HPLC samples were run undiluted with a 50-μL injection volume.

A total of 15 samples were analyzed in the comparison study (Figure 6). Titer values measured by both techniques showed excellent comparability, falling within 12% of each other, with most samples within 10%. Analysis time for the 15 samples on the Octet system was less than 10 minutes whereas the HPLC analysis required about 180 minutes.

Comparison study performed between Octet system and HPLC for Protein 1

Gold lines represent variation

Sample ID ProG HPLC (µg/mL) Octet system (µg/mL) % Difference
S1 357 318 11%
S2 2672 2395 10%
S3 388 373 4%
S4 421 448 -6%
S5 2397 2254 6%
S6 1275 1266 1%
S7 1299 1235 5%
S8 1493 1313 12%
S9 2645 2521 5%
S10 1555 1578 -1%
S11 2327 2104 10%
S12 2009 1883 6%%
S13 710 708 0%
S14 551 545 1%
S15 3233 2971 8%

Figure 6: A total of 15 samples were analyzed for the comparison study. Reported concentration values showed good agreement between the two methods, falling within 12% of each other for all samples and within 10% for most samples.

Octet Assay Demonstrated Excellent Precision

Intermediate assay precision was evaluated by comparing titer values of Fc-fusion protein control sample run on various days by several analysts, as shown in Figure 7. The overall % CV across all runs (n=54) was <5%, verifying excellent precision of the Octet assay.

Intermediate precision of the Octet assay

Figure 7: Intermediate precision of the Octet assay had an average CV of <5% (n=54).

Conclusion

The titer assay was developed at Biogen IDEC for Fc-fusion proteins in cell culture supernatants. The assay is currently used to quickly screen large numbers of clones and identify promising, high-producing clones during cell line development. More than 40,000 samples were run on the Octet QK384 system in one year, making it an indispensable tool in their cell line development workflow. Expanded use of the Octet system in other bioprocess applications is currently under exploration in the cell line development lab at Biogen IDEC.

Download PDF