Case Study: Qualifying a Commercially Available ELISA for HCP Quantification and Its Ability to Inform Purification Process Decisions

Host cell proteins (HCPs) are process-related impurities introduced into the manufacturing process by the host organism in which a biologic is produced. These proteins have the potential to elicit unwanted immunological responses in patients, which may affect the safety of the final drug product. Furthermore, residual host cell proteins, such as proteases, have the potential to degrade the biologic drug substance (DS).¹ As a result, regulatory authorities expect that HCPs in the DS will be evaluated and the most common method for HCP measurement is a quantitative enzyme-linked immunosorbent assay (ELISA). In addition to quantitating HCPs in the final DS, the HCP ELISA can also be used to quantitate HCPs in in-process samples to evaluate process performance, and to inform decisions during purification process development. It is generally accepted that a commercially available HCP ELISA can be used during preclinical and early stages of clinical development before a more sensitive process-specific ELISA, with greater antibody coverage for the antigen, is developed.¹ In this case study, a commercially available Sf9 HCP ELISA was evaluated and optimized for use in quantitating HCPs in a recombinant protein DS and in-process samples. Subsequently, information from the commercially available ELISA was used to characterize and guide downstream purification process improvement decisions.

HCP ELISAs are complex because they utilize a polyclonal anti-HCP antibody reagent to detect multiple HCP analytes. Therefore, even a commercially available HCP ELISA kit that has been qualified by the manufacturer should undergo further qualification by the user to demonstrate that the assay is capable of accurately detecting HCPs present in all of the sample types for which it is intended. In addition to the standard practice of determining the range, repeatability, intermediate precision, limit of detection, and limit of quantitation of the assay, there are particular qualification parameters which require special attention when developing a multiple-analyte HCP ELISA, including specificity, accuracy, and dilutional linearity. The vendor-supplied antibody must be shown to be specific for HCPs and not recognize (cross-react with) any DS proteins. To determine the specificity of the anti-HCP antibody provided in the kit, cross-reactivity of the antibody for DS is evaluated by 1D Western blot. A blot in which DS proteins are probed with the kit-supplied antibody and a negative result, or lack of recognition of the DS proteins by the kit-supplied antibody, is desired. Qualification of a commercially available ELISA for HCP detection should also include an assessment of the accuracy of the assay in detecting HCPs in each of the sample types to be tested, primarily to ensure a lack of sample matrix interference. To determine accuracy, spike and recovery experiments are conducted, in which the process-specific HCP standard is added to appropriate samples and a spike recovery of ±30% is demonstrated over the range of the assay.¹

In order to accurately quantitate the HCPs in a given sample, all of the anti-HCP antibodies must be in excess of their respective antigens. Determination of antibody excess is important because at high levels, a particular analyte can saturate the available antibody, leading to an underestimation of that analyte by the assay. To determine whether each anti-HCP antibody is in excess of its antigen, the property of dilutional linearity is examined. Dilutional linearity is demonstrated when a sample is serially diluted two-fold for several points and the calculated HCP concentration does not vary by more than ±20%, so that the value for the HCP concentration in that sample remains relatively constant over several dilutions.² In contrast, if the antibody is saturated by the analyte, the apparent HCP concentration in the sample will increase as the sample is diluted, leading to inaccurate quantitation of HCPs in that sample.

In addition to establishing dilutional linearity for each sample type to ensure accurate quantitation of HCPs, dilutional linearity can also be used to evaluate the clearance of HCPs by downstream process steps. Dilutional linearity indicates that each antibody is in excess of its antigen and therefore, a loss of dilutional linearity at any step in the purification process indicates that one or more HCPs is being enriched by that step, so that one or more antibodies are no longer in excess of their antigen.

To evaluate dilutional linearity, the ratio of HCP (ng/mL) to the product protein concentration (mg/mL) is plotted against the concentration of the product (mg/mL). Typically, in-process and DS samples are tested by HCP ELISA and dilutional linearity is established by demonstrating that the HCP-to-product ratio remains relatively constant regardless of the sample dilution. However, if dilutional linearity is lost during the purification process, one or more antigens are co-purifying with the product molecule and may even be enriched (assuming matrix interference and antibody-product cross reactivity are not factors).

In this case study, the effect of varying the elution conditions of a mixed mode (MM) chromatography column on HCP clearance was studied using a commercially available HCP ELISA. Several changes were made to the downstream purification process for a recombinant protein, originally designated Process A, then Process B, and ultimately Process C, as shown in Figure 1. The focus of this communication is on the MM chromatography step and the optimization that was performed on this step to improve the clearance of HCPs by this column. Although an MM chromatography step is used in all three processes, the mode of elution from the column is varied. In Process A, the elution from the MM column is performed using a phosphate gradient. In Processes B and C, elution from the column employs a pH change. Process B uses a one-step phosphate wash prior to elution by pH shift, whereas Process C incorporates a two-step phosphate wash prior to elution by pH shift. The commercially available ELISA was used to evaluate HCP levels in the column loads and eluates in each process, and ultimately guided the selection of the MM column elution conditions for optimal HCP clearance.

Figure 1. Overview of downstream purification Processes A, B and C. Elution from the MM chromatography column in Process A is by phosphate gradient. Elution from the MM column in Processes B and C is by pH change. Elution by pH is preceded by a 1-step phosphate wash in Process B and a 2-step phosphate wash in Process C.

Dilutional linearity was examined to determine the effect of changing the mode of elution from the MM column. Plotting the HCP to product protein ratio (ng/mg) against the concentration of the product (mg/ mL), shows that in Process A, dilutional linearity is maintained for each process intermediate because the Sf9 protein ratio remains relatively unchanged (measured value within 20% of previous dilution) at every sample dilution (Figure 2). In contrast, dilutional linearity was not maintained for Process B (Figure 3). In Process B, the first three inprocess samples demonstrate dilutional linearity, indicating that every anti-HCP antibody is in excess of its antigen in that sample. However, for both the MM and anion exchange (AEX) eluate samples, dilutional linearity was not observed, indicating that not all of the anti-HCP antibodies are in excess of their antigens. Since the loss of dilutional linearity occurs after changing the mode of elution from the MM chromatography column, and because dilutional non-linearity is first evident in the MM eluate sample, the data suggests that the one-step phosphate wash prior to elution by pH shift is increasing the level of HCPs co-purifying with the DS. This knowledge was used to inform the decision to include a second phosphate wash of the column prior to elution by pH shift, leading to Process C (Figure 4). Dilutional linearity was achieved in Process C, demonstrating that inclusion of a second phosphate wash prior to pH elution prevented enrichment of HCPs in the MM eluent.

Figure 2. Dilutional linearity determination by Sf9 HCP ELISA of Process A purification intermediates. The Sf9 protein ratio (ng/mg) is calculated by comparing the Sf9 HCP result (ng/ mL) to the product protein concentration (mg/mL) in a sample. Process pools from various purification steps were tested with serial dilution and plotted for Sf9 protein ratio (ng/mg) and product concentration (mg/mL). SIB: solubilized inclusion body, AEX: anion exchange chromatography, MM: mixed-mode chromatography, HIC: hydrophobic interaction chromatography

Figure 3. Dilutional linearity by Sf9 HCP ELISA of Process B purification intermediates. The Sf9 protein ratio (ng/mg) is calculated by comparing the Sf9 HCP measurement (ng/mL) to the product protein concentration (mg/mL) in a sample. Process pools from various purification steps were tested with serial dilution and plotted for Sf9 protein ratio (ng/mg) and product concentration (mg/mL). SIB: solubilized inclusion body, CEX: cation exchange chromatography, MM: mixed-mode chromatography, AEX: anion exchange chromatography

Figure 4. Dilutional linearity by Sf9 HCP ELISA when eluting from the MM chromatography with pH following a 2-step phosphate wash of the column. The Sf9 protein ratio (ng/mg) is calculated by comparing the Sf9 HCP measurement (ng/mL) to the product protein concentration (mg/mL) in a sample. Process pools from various purification steps were tested with serial dilution and plotted for Sf9 protein ratio (ng/mg) and product concentration (mg/mL). SIB: solubilized inclusion body, CEX: cation exchange chromatography, MM: mixedmode chromatography

The ability to elute the drug substance from the MM chromatography with pH instead of phosphate was an important change to the downstream purification process that resulted in an overall improvement in the removal of HCPs by the MM column and reduction of HCP in the DS. When eluting by phosphate, there was not a significant reduction in HCP levels by the MM column in Process A (Figure 5A). In contrast, the same MM purification step in Process C (Figure 5B) where the mode of elution has been changed to pH with a two-step phosphate wash prior to elution reduced HCP levels by two orders of magnitude. Therefore, the optimization of the MM chromatography step contributes greatly to the overall lower levels of HCPs in the MM eluate and DS.

Figure 5. Optimization of the MM chromatography wash step reduces overall HCP levels in the MM eluate and resulting DS. In Process A (Panel A), there is less efficient clearance of HCPs by the MM column using phosphate elution compared to Process C (Panel B), resulting in higher HCP levels in the DS. The Sf9 protein ratio (ng/mg) is calculated by comparing the Sf9 HCP level (ng/mL) to the product protein concentration (mg/mL) in a sample. Process pools from various purification steps were tested and the Sf9 HCP-toproduct ratio (ng/mg) was plotted for each. SIB: solubilized inclusion body, AEX: anion exchange chromatography, MM: mixed-mode chromatography, HIC: hydrophobic interaction chromatography, CEX: cation exchange chromatography, UF/ DF: ultrafiltration/diafiltration

Because residual host cell proteins in a biologic can have many unwanted effects, including a reduction in safety and efficacy of the therapeutic, the level of host cell proteins should be quantitated and minimized.¹ The use of a commercially available Sf9 HCP ELISA is a powerful tool for quantitating HCPs in process-specific samples and evaluating downstream purification process changes. Due to the complex nature of a multi-analyte ELISA, it is imperative to qualify a commercially available HCP assay for use in quantitating HCPs in process-specific samples. It is especially important in the qualification of HCP ELISAs to determine specificity, accuracy, and dilutional linearity to demonstrate that the commercial ELISA is capable of accurately quantitating HCPs in each type of downstream process sample. Once the assay is qualified for testing process-specific samples, the results can be used to evaluate downstream purification process changes and optimize conditions to reduce overall HCP levels. In this case, the data demonstrate how quantitation of HCPs by the commercial ELISA was vital for characterizing downstream process changes. Even small alterations to the process, such as changing column wash conditions, can have a significant impact on the overall clearance of HCPs by the MM purification chromatography. In particular, examination of the dilutional linearity is a powerful and specific method important for evaluating downstream purification changes on HCP clearance and optimizing the MM chromatography column wash conditions to avoid enrichment of HCPs. In this study, the characterization of downstream process changes not only led to improved clearance of HCPs by the MM chromatography, but also resulted in overall lower levels of HCPs in the purified DS.

AMA References:

1. Residual Host Cell Protein Measurement in Biopharmaceuticals (USP 39-NF 34). Vol 1. Rockville, MD: United States Pharmacopeia Convention; 2016: 1416-1436.
2. Assay Qualification Template for Host Cell Protein ELISA. Cygnus Technologies. Available at: https://cygnustechnologies.com/content/faq.html#1. Accessed August 28, 2017.

Author Biographies