Keeping Tabs on Host-Cell Proteins


Protein-based biologics for therapeutic use represent a rapidly expanding class of biopharmaceuticals. Dominating this category are the monoclonal-antibody therapeutics, sharing an estimated market worth of almost $80 billion along with antibody diagnostics, as of 2015 [1]. This phenomenal share of the market is largely thanks to the clinical successes demonstrated by pioneering antibody therapeutics, such as the promising immune-targeting therapies for the treatment of cancer that have emerged recently. Although comprising a smaller portion of the larger biotherapeutics market, peptide-based therapeutics represent a growing sector, as advances in technology have led to more stable products, with almost limitless modification potential.

A large proportion of antibody- and peptide-based therapeutic production now relies on in vitro expression. Currently, the majority of these comprise mammalian-cell-based systems. Jonathan Royce, global product manager for antibody affinity media at GE Healthcare’s Life Sciences business, explains that it comes down to molecular complexity. “Mammalian cells have all of the mechanisms required to properly express, build and glycosylate antibodies,” states Royce. However, peptide therapeutics, such as Avacta Life Sciences’ Affimer® Proteins, are much simpler molecules than antibodies, so a growing number are being produced in bacterial hosts.

Whatever the means of production, when it comes to therapeutics, safety is paramount. 

Quantitative analysis

Being able to remove host-cell impurities is one thing, but being able to detect and quantitate HCPs, and then demonstrate to what extent they have been removed, is a key requirement of drug development. Detection of HCPs is most often performed using an ELISA, but this can also be done by Western blotting. The basic premise of such tests is to use a polyclonal antibody pool, raised to complement the host cell’s protein profile, to capture and quantify the total amount of HCP present in a given sample.

For preclinical and clinical development stages, off-the-shelf HCP ELISA test kits can be used. There are specific kits available based on the host-cell system in use. Enzo® Life Sciences, for instance, provides both E. coli-specific tests and mammalian, Chinese hamster ovary (CHO)-targeted ones. Other examples of companies providing these test kits are Cisbio Bioassays, Crystal Chem, Cygnus Technologies and FortéBio.

However, as the production process advances to scale, its detection requirements start to change as regulatory demands begin to apply. Testing must become ‘process-specific’ to meet regulatory specifications. This means the polyclonal-antibody complement used for detection must represent the exact host system being used.

“When you transform cells, you can get slight genetic variations that may result in a slightly different profile of host-cell contaminants,” cautions Amrik Basran, chief scientific officer at Avacta Life Sciences. “Even changing temperature in the fermenter or changing oxygen rate can change a lot of growth parameters in the vessel and therefore the contamination profile of your batch,” Basran states. This is why an individual process requires a bespoke antibody pool matching the exact contamination profiles obtained from it. “Because then you know that you’ve got the exact polyclonal-antibody profile for the impurities made during your process,” adds Basran.

Michael Brewer, director of pharmaceutical analytics at Thermo Fisher Scientific, manages a team that provides fully integrated systems for quality and safety testing of pharmaceuticals during the manufacturing process. Brewer’s team also develops technologies to move HCP testing forward. “Since biotherapeutics emerged in the late 1970s, ELISAs have remained the core HCP assay; however, in the last few years, we’ve started to work on technologies to replace these,” shares Brewer. These newer technologies still use antibody pools for detection, but utilize different methods to obtain their results. Thermo Fisher Scientific’s ProteinSEQ™ assay, for example, uses an antibody pool attached to magnetic beads to capture HCPs. Instead of detection with an enzyme-linked secondary antibody, the assay relies on qPCR—the detection antibodies are linked to a sequence of DNA that can only be amplified after antigen binding takes place. Thus, this method combines the specificity of antibody-antigen interactions with the range and sensitivity of PCR. “Additionally, the qPCR-based readout generates a standard curve that is quite a lot more linear than a typical ELISA assay,” according to Brewer.

Thermo Fisher Scientific is not the only company providing solutions for HCP monitoring and detection at higher levels of sensitivity; Gyros’ Gyrolab™ offers an updated take on the traditional ELISA system, and FortéBio’s Octet brings a biosensor-based system to the market. No matter the method of choice, Brewers points out that as long as drug manufacturers can demonstrate HCP clearance below the level of the current specifications, their products should meet less resistance when it comes to gaining regulatory approval.

In the end, there will always be trace elements of HCPs, no matter how rigorous the production process. “We’re talking about measuring things down below the level of detection or down below a limit someone has decided is acceptable,” says Royce. It may also be the case that regulators ask for characterization of the HCP component of a process. Here, the downside of ELISA testing is that it can’t provide information on the specific contents of the HCP profile. “You get a general quantification of how much is there, but you don’t know the exact nature of the species that you’re dealing with,” states Royce. In this case HPLC, or even UPLC combined with mass spectrometry, is required “if you really want to understand the complex mixture of host-cell proteins that you’re dealing with,” Royce says.

Stability testing

In addition to HCPs, the therapeutic antibody or protein itself may pose a risk of ADRs or be rendered ineffective if drug selection and stability testing aren’t smart enough.  

“When people think of purity, there’s obviously consideration of the contaminant profile, but there’s also the purity and stability of the molecule itself,” states Basran.

“Certainly, aggregation is one of the key indications of stability,” Brewer says, adding glycosylation patterns to the list. Royce adds that “developers are also looking at removing things like certain charge variants of antibodies.”

When it comes to later-stage process development or formulation procedures, immunological assays may not provide enough detailed information about the target molecule. For instance, you may take a sample you have tested routinely with ELISA and find its IC50 unchanged, but then identify 10% aggregation when you transfer that same sample to a size exclusion column (SEC), Basran explains.

Technologies like SEC and analytical ultracentrifugation generate better-quality data when examining things like aggregation. But even in this respect, scientists are looking for higher throughput (HTP) methods, “especially at the bigger biopharma companies—where thousands of samples need to be tested in a given year,” Brewer says. Capillary electrophoresis (CE) methods are an alternative option, enabling the simultaneous testing of multiple samples compared with, say, UPLC, which only accommodates a single sample at a time. “We’ve launched a product for characterization of glycosylation patterns that’s based on CE technology,” states Brewer, referring to the GlycanAssure™ glycan analysis system, a multicapillary CE instrument-based system that offers high-resolution glycan analysis of up to 24 samples at once.

Whatever options researchers chose to interrogate the stability of their therapeutic protein, they need to make sure they use an orthogonal approach—and not just at the end of the process. “As early as possible, you also need to think about the developability of your therapeutic,” advises Basran.

“You’re not just trying to identify the lead binder. You need to think about stability and scalability early on, as well,” comments Royce. “People are looking at things like aggregate quantity quite a bit earlier than they probably were 10 years ago.”

For Basran, when it comes to safety testing, it’s always best to be critical of your process and seek further validation of any assay you choose. “I always say, Think about what your assay tells you, but also what it doesn’t tell you,” comments Basran. Thus, fully integrated, orthogonal approaches are key to both HCP and stability testing of both antibody- and peptide-based therapeutics.


[1] Fung, P, “2015 Antibody Market Report,” Biocompare, online article, September 8, 2015.

[2] Baker, MP, et al., “Immunogenicity of protein therapeutics: The key causes, consequences and challenges,” Self Nonself, 1(4):314-322, 2010. [PMID: 21487506

[3] Eaton, LC, “Host Cell Contaminant Protein Assay Development for Recombinant Biopharmaceuticals,” J Chromatograph, 705(1):105-14,1995. [PMID: 7620564

Image: Shutterstock Images