BEBPA 2017 – Key topics and take-home messages

16 May 2017

The BEBPA Host Cell Protein (HCP) Conference 2017 was again held in charming San Francisco with HCP experts from all over the world. 140 engaged participants from biopharma companies, instrument vendors, CRO’s and regulatory authorities had a very good HCP conference with high level presentations and an excellent forum for discussion of ELISA and orthogonal method. Kevin Van Cott chaired Pre-conference educational workshops on LC MS for HCP analysis, and Denise Krawitz chaired HCP analysis 101 with focus on immunoassay.

The BEBPA 2017 host cell protein conference addressed many important aspects of HCP analyses, including orthogonal method for bio pharmaceutical development.

For me as a protein mass spectrometrist at Alphalyse, the key topics and take-home messages were:
(Click to download as PDF document)

ELISA HCP Assays

  • ELISA is still the gold standard method for HCP measurement.

    The ELISA measurement units for the HCP content are Immuno-equivalents to “ng HCP / mg drug substance”.

    ELISA does not measure the actual content of nanograms of host cell protein, as it is based on an animal immune-response to a pool of antigens injected into a rabbit or goat. Some of the HCP antigens are detected well by high affinity antibodies in the ELISA. Some antigens with low affinity antibodies, and some antigens are not detected by the antibodies at all. Therefore, the ELISA measures “Immuno-equivalents” of the nanogram content of process-related impurities.

 —

  • The ELISA HCP analysis is a tool to monitor manufacturing process consistency with regards to HCP content.

    If deviations occur in the down stream protein purification, then the ELISA HCP assay should be able to detect the batch to batch variation in HCP content. For example if there is a leaky chromatography column. Or in case of a variation in total HCP content from cell culture media, cell extracts or bacterial derived inclusion bodies. The ELISA antibodies should have a broad coverage to detect HCP’s. Both those occurring early and those in later stages of the purification process. The presenters did not recommend to use down stream purified HCP antigens for immunization to increase coverage. The ELISA should not be developed for specific detection of the few HCPs present in the final drug substance.
    Consequently, ELISA is not a good tool to monitor the few hitchhiker HCPs in a highly purified DS.

  —

  • The quality of the ELISA reagents should be demonstrated fit-for-intended-purpose on the specific bioprocess.

    You can perform  HCP coverage analysis on the antigen pool used for immunization with the antigen standard used in the ELISA assay. It is also possible to perform coverage analysis by comparison of the ELISA antigen standard versus a null cell line without the drug substance expression. Alternatively the harvest sample of the drug substance.

  —

  • It is still common to perform the ELISA HCP coverage evaluation by 2D PAGE and 2D PAGE Western Blotting with HCP antibodies.

    However, a coverage number in % is very difficult to determine in a reproducible and trustworthy way. This is because protein smears in the 2D images and complex spot detection. Several presentations showed alternative HCP antibody coverage analysis by 1D SDS PAGE Western Blotting, and by immunoaffinity purifications in combination with 1D SDS PAGE and/or LC MS analysis.

  —

LC MS/MS HCP Analysis

  • LC-MS analysis is the most important orthogonal method to ELISA for HCP analysis.

    This is because it complements an ELISA derived number of immune-equivalents with a direct identification of peptides/proteins present in the sample. You can perform LC-MS analysis either as a direct LC-MS analysis of in-solution digested process samples or in combination with 1D SDS PAGE, 2D PAGE. The host cell proteins are digested with trypsin in-solution or in gel bands/spots. Alternatively, sensitive LC-MS/MS analyze the peptides, and a search in a protein database for the specific drug substance identifies the HCP peptides. The protein database contains sequences from the UniProt or NCBI proteome. This includes the expression organism, the drug substance protein sequence, the trypsin sequence, and frequent contaminants.

  —

  • Some of the major challenges for LC-MS for routine high throughput HCP analysis are:

    • Both the chromatography system and the MS system need a high dynamic range. This is necessary for analysis of low ppm HCP levels in the presence of large amount of drug substance protein.
    • Long analysis run time due to the use of multi-dimensional chromatography systems and nanoLC flow rates needed for sufficient separation of HCP peptides from drug substance peptides, and for high sensitivity detection of the HCPs.
    • Lack of good accurate quantification methods that works both on the complex harvest samples with hundreds of HCPs as well as the purified drug substance with low ppm levels of a few HCPs.
  • Several of the talks and posters presented possible solutions for these challenges:

    • 1-dimensional chromatography setup using Waters CSH columns. These should have high loading and peak capacity. Use formic acid in mobile phase for compatibility with MS analysis.
    • The 1D chromatography run with 45min-80min gradients at higher flowrates for robust chromatography.
    • Quantification using label-free methods including spike-in of known amounts of protein standards, for example the Universal Proteomics Standard (UPSII, Sigma). Use MRM methods and stabile isotope labeled peptides for sensitive quantification of specific low level HCPs.

  —

  • SWATH LC MS analysis for HCP identification and quantification.

    Three presentations from Abbvie, Alphalyse & Savara, and BioProcessing Technology Institute showed very promising results on data-independent acquisition of MS data using Sciex mass spectrometers. The advantage of data-independent acquisition includes high reproducibility in protein identifications. Other advantages include a high sensitivity for low level HCPs, and accurate quantification using Hi50 and Hi3 peptide signals. Examples of dilution linearity of spiked standard proteins demonstrates the power for quantification of proteins identified with this method. This is also true for  dilution linearity of each individual HCP.

  —

  • Regulatory aspects of HCP impurity specifications.

    Regulatory experts from Paul-Erlich-Institut, US-FDA, and Health Canada explained the regulatory framework for setting HCP specifications and control. The presentations featured several case examples. The presenters advocated future implementation of orthogonal methods to the standard ELISA assays. They can identify and evaluate the risk of the host cell proteins present in drug substance used for clinical applications. Furthermore, the presenters strongly suggested registration applicants to seek regulatory advice early in development. Do this for justification of the suggested HCP analysis strategy and also to provide reliable data to support the strategy.

  —

Overall the BEBPA 2017 HCP Conference was an excellent meeting facilitating high level discussions. In particular by increasing our understanding of this emerging field and push improvements in orthogonal method & technologies. We are looking forward to contributing to the BEBPA Host Cell Protein conference in 2018.

Yours Sincerely

Ejvind Mørtz, COO, Alphalyse