How To Mitigate The Risks Posed By "High‐Risk" Host Cell Proteins
By BioPhorum
Host cell proteins (HCPs) are process‐related impurities that may copurify with biopharmaceutical drug products. Some of these are problematic and can be considered high‐risk and can include those that are immunogenic, biologically active, or enzymatically active with the potential to degrade either product molecules or excipients used in formulation. Some have been shown to be difficult to remove by purification.
There are only very few known cases where an HCP has directly impacted a patient’s safety, so why should the biopharmaceutical industry worry about these high‐risk HCPs? What approach could be taken to understand the origin of their copurification and address these high‐risk HCPs?
To answer these questions, the BioPhorum Development Group (BPDG) HCP Workstream initiated a collaboration among its 26‐company team to create industry alignment around high‐risk HCPs. The result is a full-referenced paper called “High‐risk” host cell proteins (HCPs): A multi‐company collaborative view, which is summarized in this article.
The HCP Workstream has built on previous cross‐industry groups that have made recommendations for assessing HCP impurities and developing a common understanding of regulatory agency requirements.
A subteam also performed literature searches to understand the impact of high‐risk HCPs and discuss member companies’ experiences. These focused on biopharmaceuticals produced in Chinese hamster ovary (CHO) cells and purified by a Protein A affinity column and additional polishing steps.
Frequently Seen HCPs In Downstream Processing
The CHO expression system is one of the most common mammalian systems in biopharmaceuticals production due to its ability to produce complex proteins with post-translational modifications similar to those produced in humans. More than 6,000 CHO HCPs have been identified.
Downstream processing often includes Protein A affinity purification followed by additional polishing steps to further remove aggregates, charge variants, HCPs, and host cell DNA. The literature search performed on the HCPs present after Protein A chromatography revealed that many of the same HCPs are found across the biopharmaceutical industry. The HCPs identified were similar despite differences in CHO cell lines, upstream processes, or downstream processes.
The team compiled a list of these frequently seen HCPs from published literature, which can be used to reference HCPs found throughout different processing steps. These are fully listed in a table in the BioPhorum document, which provides the molecular weight, pI, the number of amino acids, and a direct link to the Uniprot database for each protein. See Table 1 for an extract.
Table 1: A comprehensive list of frequently seen HCPs found throughout different processing steps (extract only)
CHO protein |
Molecular weight (kDa) |
pl |
UniProt accession no. |
No. of amino acids |
40S ribosomal protein SA (RPSA) |
19.7 |
9.4 |
G3HQX0 |
179 |
60S acidic ribosomal protein P0 |
30.6 |
8.9 |
G3HKG9 |
280 |
78 kDa glucose regulated protein (GRP78, BiP) |
72.4 |
5.1 |
G3I8R9 |
654 |
Actin, cytoplasmic 1 (ACTB) |
41.7 |
5.2 |
P48975 |
375 |
Alpha‐enolase (2‐phospho‐D‐glycerate hydro-lyase) |
15.5 |
5.0 |
G3I0W1 |
139 |
BioPhorum’s HCP searchable database will continue to be updated with frequently seen HCPs and their characteristics (see Table 2). The identities of the different HCPs may be useful for companies as a guide when developing a downstream process or possibly refining existing purification platforms.
Table 2: Database extract of frequently seen HCPs – cathepsin CHOs only
CHO protein |
UniProt accession no. |
Molecular weight (kDa) |
pl |
No. of amino acids |
Cathepsin B (CatB) |
37.5 |
5.7 |
339 |
|
Cathepsin D (CatD) |
44.1 |
6.5 |
408 |
|
Cathepsin E (CatE) |
42.2 |
4.7 |
388 |
|
Cathepsin L (CatL) |
37.3 |
6.8 |
333 |
|
Cathepsin Z (CatZ) |
34 |
7.5 |
306 |
Classification Of HCPs
The classification of a problematic HCP as high or low risk for immunogenicity can be difficult. The risk depends on many factors, including the drug indication, route of administration, frequency of administration, and the amount of HCP per drug dose administered.
An approach for classifying problematic HCPs as high‐risk could be based on an HCP’s ability to copurify with the product, the frequency at which it is seen in downstream processing, its ability to modify or degrade the drug and/or the excipient, and its potential for immunogenicity.
Using this approach, HCPs can be classified based on their potential impact into four major categories: product quality, formulation, direct biological function in humans, and immunogenicity. Through extensive literature searches and working experiences of the HCP Workstream, a list of high‐risk HCPs was compiled and categorized based on their impact (these are listed in a table in the BioPhorum publication and within the database).
HCP Workstream Survey Results
After thorough literature searches and BPDG team discussions, the team was surveyed to understand the collective experience of the member companies with problematic or high‐risk HCPs. A summary of the key findings is:
- 69% of companies indicated that they had experienced issues with individual HCPs during drug production.
- Phospholipase B‐like 2 was the HCP with the highest response for identification/detection and clearance after detection. Other HCPs identified included lipoprotein lipase, cathepsins (D, B, and L), and lysosomal phospholipase A2. Most respondents indicated that they were able to show clearance of these HCPs.
- Most companies used mass spectrometry (MS) or enzyme‐linked immunosorbent assays (ELISAs) specific to the HCP to identify and measure individual HCPs. Fewer respondents used enzyme/activity assays and only one indicated the use of gel excision followed by liquid chromatography‐mass spectrometry.
- Most respondents used one or more HCP ELISA method for release testing to quantify the total HCP in drug substance, and a few used technologies such as MS quantitation, gyros, and meso scale discovery.
- Most respondents used MS for process development support; others used it for clinical and nonclinical drug substance/drug product samples. Also, 64% of companies used relative quantification, the rest used absolute quantification.
- 67% of respondents did not enrich for HCPs before analysis, while 17% did enrich and the rest used precipitation. For digestion methods, 13 companies used trypsin, while four used trypsin + lysine C.
- 45% of companies said they did not receive regulatory feedback about their analytical testing strategy for total HCP and individual HCPs; 36% said they did. The rest said that 2D coverage was requested as part of regulatory feedback.
Recommendations For Industry
Establishing a link between the identified HCP and its impact on the process or the patient can be challenging. Therefore, industry must create a comprehensive analytical strategy to properly measure known HCP impurities and, if identified as high‐risk, develop a control strategy to monitor and/or eliminate these impurities. The recommendations to achieve these aims are:
- The critical reagents needed to develop the total HCP immunoassay should be carefully generated and characterized.
- The critical antibody reagents for the process-specific HCPs should be assessed using additional orthogonal methods such as 2D‐gels/westerns and/or antibody affinity extraction followed by MS.
- The total HCP ELISA should be routinely performed to determine the levels of HCPs in the process samples.
- Orthogonal methods and/or extra individual HCP detection methods are recommended to ensure product purity.
- If an individual HCP is identified and considered to be potentially high‐risk:
- A specific immunoassay to monitor the known impurity or a targeted MS analysis method may need to be developed to better understand the level of HCP and its potential impact.
- If the identified impurity has enzymatic activity, an activity assay could be developed to guide process and product development to monitor, remove, and/or inactivate the protein.
- An immunogenicity assessment could be performed through an in-silico method, followed by in-vitro comparative immunogenicity assessments if in-silico prediction indicates a high immunogenicity risk.
- Toxicology input could be valuable in assessing the clinical risk.
HCP detection, quantitation, and removal from the final biotherapeutic process can be complex. BioPhorum’s list of high‐risk or problematic HCPs can be a resource for companies developing biotherapeutics in CHO cells. It classifies them into four major categories based on their impact on product quality, formulation, direct biological function in humans, and immunogenicity (see Table 3). If an HCP on the high‐risk list was identified in a process, companies should follow the recommended actions, perform a risk assessment, and establish an HCP control strategy.
Table 3: High‐risk HCPs categorized based on their impact (extract only)
Protein name |
Function |
Impact |
Type of impact |
Select references |
78 kDa glucose regulated protein (GRP78; BiP) |
Protein folding and quality control in the endoplasmic reticulum lumen |
Drug quality |
Aggregation of drugs |
Farrell et al., 2015; Liu et al., 2019; Valente et al., 2015 |
Alpha‐enolase |
Catalyzes the dehydration of 2‐phosphoglycerate to phosphoenolpyruvate |
Drug quality |
Modification of drug |
Valente et al., 2015; Zhang et al., 2014 |
Annexin A5 (ANXA5) |
Binds with high affinity to phospholipids and serves as a marker for apoptosis |
Immunogenicity |
Immunogenic response |
Fukuda et al., 2019; Gilgunn & Bones, 2018 |
C‐X‐C motif chemokine 3 (CXCL3) |
Cytokine with potential oncogenic properties |
Biological function in humans |
Immunogenic response |
Gilgunn & Bones, 2018; Gilgunn et al., 2019 |
Carboxylesterase (CEB) |
Catalyzes the cleavage of ester‐ or amide‐containing substrates into alcohol and carboxylic acid |
Formulation |
Degradation of polysorbates |
McShan et al., 2016; Zhang et al., 2020 |
Each company may deal with unique circumstances during their product and process development and must perform the HCP risk assessment on a case‐by‐case basis. We hope this collaboration will guide industry on targeted high‐risk HCP characterization and proactively mitigate the risks posed by these HCPs in biological products.