Olaf Stamm, PhD

In AAV Gene Therapy, The Road to Purification is Rocky

So, trust but verify. Mass Spectrometry can Support Generic ELISAs in detecting host cell proteins.

Residual Host Cell Proteins (HCP) are inevitable impurities in any biopharmaceutical product. In contrast to traditional recombinant proteins like monoclonal antibodies manufactured in Chinese Hamster Ovary (CHO) cells, AAV-based gene therapy products present additional HCP control challenges from a strategic process development perspective. AAVs, or adeno-associated viruses, are typically manufactured using human cell lines like HEK293, which present a potentially complex proteome background. Contrasting with recombinant proteins purified from the cell culture supernatant, AAV harvest typically relies on cell lysis, meaning the entire proteome may contribute to the HCP load presented to the downstream purification process.

Additionally, instead of the chemically defined media used in CHO production processes, virus production relies on protein growth factors and protein-containing additives including serum originated from other non-host cell species. Finally, the purification steps used for viral vectors are limited in number, impacting their overall capacity to separate the AAV product from other proteinaceous impurities.

Adding to the technical challenges, gene therapy programs are often intended for severe indications and are therefore more likely to be placed on an accelerated development pathway. This time pressure conflicts with the timeline required to develop a product-specific immunoassay, so commercially available HCP detection ELISA kit use is common practice from early stage through late stage AAV product development. This approach poses significant risk; the generic HCP antigen used to generate antibodies for a generic ELISA kit may not provide accurate HCP quantitation since the actual HCP antigen in an AAV product will de-facto be different.

As shown below,  after comparing HCP content in drug substance samples of AAV vectors measured using a generic kit-based ELISA, the actual content determined by quantitative LC-MS showed consistently higher HCP values across many lots than were observed using generic ELISA, indicating commercial HCP kits may in some cases significantly underestimate HCP content by a wide margin. The impact of this is especially relevant to high-dose AAV programs. The root cause of these findings can be explained as insufficient coverage by the antibodies used in the ELISA, as demonstrated by 2D SDS-PAGE and Western Blot.

Slide compares Elisa with mass spectrometry in assessing HCPs in C&GTs

The HCP content is a critical quality control attribute highly relevant to the safety of a gene therapy product. Therefore, the analytical methods used in process development and quality control should be suitable for their intended use by providing accurate results. In cases where the development timelines are forcing use of generic reagents, those reagents should be checked for specificity and coverage (accuracy) and the results obtained be confirmed by an unbiased approach such as Mass Spectrometry.

Quantification by LC-MS is more accurate because known protein/peptide standards can be spiked into the test articles providing a reliable quantitative readout of actual HCP content. Furthermore, LC-MS provides identification of the individual proteins detected, guiding refinement of AAV purification strategies on a faster timescale than product-specific ELISA development.

Here are eight high-risk mitigation strategies for assessing your clinical products.

• Don´t trust ELISA kit results blindly. Remember that no single commercial ELISA is fully capable of detecting such a heterogeneous mixture.

• Check coverage of kit antibodies (at least for linear coverage by 2D SDS-PAGE/Western Blot). Since every process is unique, no single antibody works well for all processes. Make sure you choose the antibody that works best for your process.

• Verify results from immunoassays by orthogonal tools (Mass Spectrometry) at least on samples from drug substance, ideally from DSP intermediates, too. Our findings found that the ELISA kit underestimates host cell protein content in Drug substance by an average factor of 33.

• Check for potential detection gaps of the immunoassay. While immunoassays provide an aggregate measurement of many individual proteins, these measurements can also be misleading due to multiple factors. Tandemv Mass Spectrometry may help alleviate this problem, however.

• For investigational medicinal products, If required, close detection gaps by peptide immunization. Pooling of antibodies from different immunizations can help achieve this.

• Consider developing a project specific assay - early! Remember, if a gene therapy product is delayed in getting regulatory approval the setback could hold up drug development long enough for a competitor to get their product to market first.

• Leverage proteomics results from early phase into project specific assay development. Remember that different cell lines show different and complex proteomes and secretomes.

• Consider the biology of the expression system: Is the product secreted or intracellular, does the expression system require additional steps/reagent for breaking up the cell, does the expression system get changed by the product expression (virus expression does impact proteomics of a cell line).

• Go for a holistic approach – immunoassay and mass spectrometry. Results matter. By combining forces you can ensure your findings are accurate, which will help strengthen your product’s profile.