Downstream processing of biopharmaceuticals is costly and time-consuming, often involving multiple steps with significant time and energy expended on maximizing product quality and yield. Affinity chromatography is one of the simplest and most effective methods for purifying protein and peptide therapeutics, offering reduced process steps and therefore higher yields than nonaffinity methods can provide. Protein A is widely used for affinity purification of monoclonal antibodies (MAbs), Fc fragments, and Fc fusion proteins. But it is a challenge to find a protein A equivalent for nonantibody targets.

To be useful in bioprocess applications, an affinity medium must provide the specificity required to purify a product effectively while allowing elution conditions that preserve product structure and activity. It is also important that an affinity medium can withstand cleaning processes required for effective adsorbent cleaning to meet regulatory standards in therapeutics production.

To address these challenges, BAC derived a class of affinity ligands from unique heavy-chain antibodies found in genus Camelidae (trade named CaptureSelect). Heavy-chain antibodies are devoid of the entire light chain and CH1 domain found in conventional antibodies (Figure 1). Because of that absence, antigens are bound only by the variable domain of the heavy chain (VHH) without any less binding affinity than conventional antibodies. Derivative VHH ligands are ~12-kD single-domain fragments that make up the antigen-binding domain. They are efficiently produced by the yeast Saccharomyces cerevisiae in a system that is completely free of animal components. This new class of ligand has been shown to provide not only a suitable alternative to protein A (1), but also an equivalent for a wide range of nonantibody targets, a number of which are being used in manufacturing clinical trial materials (CTMs).

PRODUCT FOCUS: NON-MAb biologics
PROCESS FOCUS: Downstream processing
WHO SHOULD READ: Process development and manufacturing
KEYWORDS: Chromatography media, platform technologies, purification, specificity
LEVEL: Intermediate

The unique suitability of the VHH domain for use as an affinity ligand lies in the stability of its structure as a fragment. It is the smallest known functional antigen-binding antibody fragment that shows high affinity and stability. The Fv domain is its equivalent fragment in conventional antibodies that consists of a heavy-chain variable domain (VH) and a light-chain variable domain (VL) held together by a noncovalent bond. If that bond is broken, all antigen-binding functionality is lost. Although the method of synthetic production — linking the two variable domains together into a single chain — retains the antigen binding functionality of scFv fragments, they often show less affinity than their parent antibody. Because the VHH domain exists as a single polypeptide chain, it is an extremely stable and easily produced fragment that retains the full binding activity of its parent heavy-chain antibody.

VHH fragments have been used successfully as affinity ligands in chromatography processes for purifying many types of protein, including antibodies (2,3,4,5), blood factors (6), and adenoassociated viruses (7,8). Ligands for a specific target molecule are identified from expression libraries that represent the heavy-chain antibody repertoire of an immunized llama. Coming from a highly variable mammalian immune system, these expression libraries provide an opportunity to select VHH molecules that meet very specific requirements for specificity, affinity, elution profile, and cleaning conditions. Thus, VHH affinity ligands can be created to match exactly the processes for which they are intended. Immobilization onto resins suitable for large-scale bioprocess applications make these ligands good tools for downstream processing.

Platform Technology for Multiproduct Facilities

VHH affinity media enable a “plug-and-play” platform approach providing a highly specific primary capture step that can be slotted into existing downstream processes. Figure 2 illustrates how VHH media can be incorporated into any “generic” biopharmaceutical purification process for comparable results with current MAb purification platforms using protein A as a primary capture step.

Many biotherapeutics are produced in multiproduct contract manufacturing facilities. It would be significantly convenient if purification processes for most/all products could be made similar — for example, if nonantibody products could be purified by a similar process to the protein A purification of MAbs. To that aim, the use of a VHH ligand for primary capture of recombinant Factor VIII (FVIII) was investigated as an alternative to a traditional multistep process involving chromatographic separation and precipitation techniques.

The VHH adsorbent was found to be highly selective for recombinant FVIII. It provided good viral clearance and significantly improved host-cell DNA and protein clearance (6). Of importance, FVIII elution was achieved at a neutral pH because the protein is relatively unstable, and yields exceeded 70%. In addition, the adsorbent retained capacity following 25 cycles of cleaning and reuse, and it produced consistent results across the range from laboratory to manufacturing scale. That demonstrated the suitability of the VHH ligand for use in FVIII commercial manufacturing. Table 1 summarizes the key features of the anti-FVIII adsorbent.

Table 1: Summary of FVIII adsorbent analysis (6)

Customized Specificity

VHH ligands are selected and developed from expression libraries obtained from a llama immunized with a target molecule. Each library contains thousands of potential ligands that can be screened and tested for specificity/affinity for the target molecule. Specificity can be as narrow or as broad as required, and binding characteristics can be tuned to produce a desired elution profile. Table 2 summarizes the VHH ligands currently commercially available.

Table 2: Commercially available VHH ligands

To assess the specificity of the VHH ligand against human IgG (IgSelect, GE Healthcare), an ELISA was used to compare the distribution (by percentage) of IgG subclasses in serum between the load and elution fractions (Table 3). Subclass distribution in the loading material was retained during purification of human IgG from serum on the VHH affinity resin (anti human IgG-Fc). By comparison, because protein A–based affinity resins lack a sufficient binding affinity toward human IgG3, the ratio of subclasses in a protein A–purified pool becomes unbalanced.

Table 3: IgG subclass distribution (%) in load and elution fractions determined by ELISA

Table 3 shows how it is possible to develop a VHH ligand with a relatively broad specificity for all human IgG subclasses. However, VHH ligands have also been developed showing excellent human IgG subclass selectivity (e.g., for IgG1, IgG3, or IgG4). To assess the efficacy of a VHH ligand with that much narrower a specificity (IgG4) an ELISA was used to compare the IgG subclass distribution (by percentage) in IVIG between load and elution fractions (Table 4). Human IgG4 subclass selectivity was demonstrated by a >20-fold enrichment of IgG4 in a one-step purification.

Table 4: IgG subclass distribution (%) in load and elution fractions determined by ELISA

Human IgG–specific VHH ligands have obvious application in MAb production. The ability to select for species specificity during development of a VHH ligand has provided a powerful tool for one-step purification of therapeutic human IgG from feedstock materials that come from transgenic animals (e.g., cattle). These media enable removal of low levels of endogenous IgG antibodies that may be coexpressed with the human IgG. Table 5 demonstrates a significant clearance of bovine IgG contaminants (≥97%) from a human IgG product using IgSelect human IgG VHH resin from GE Healthcare (www.gelifesciences.com). In a separate study, the human IgG ligand was used in conjunction with a selective bovine IgG scavenging ligand to produce a human polyclonal IgG product at a purity suitable for therapeutic use (5).

Table 5: Removal of bovine IgG from human IgG samples using 5 mL IgSelect human IgG VHH resin

Matching Elution Profile to Target Requirements

Unstably structured proteins may be difficult to effectively purify using affinity chromatography because of their instability under conditions required for elution, such as low pH or the presence of high solvent concentrations. The inherent flexibility of the VHH ligand development process means that ligands can be selected for development based not only on their specificity, but also on their elution profiles. Elution under mild conditions is assessed by testing ligands with a range of mild elution buffers. Table 6 summarizes those profiles achieved for a range of target molecules purified with highly specific VHH affinity resins.

Table 6: Elution of a range of target molecules under mild conditions.

As described above, FVIII is a relatively unstable molecule. It consists of a 200-kDa heavy chain and an 80-kDa light chain held together with a highly labile noncovalent metal bond (Me²⁺). The protein is stable in neutral conditions for only a limited time and is not stable at all at low pH. The challenge was therefore to identify a ligand that would enable swift (one-step) purification of the FVIII molecule at neutral pH. The ligand identified was highly specific and selective for FVIII products. An elution buffer optimization step provided the balance of buffer components required to achieve high product recovery at neutral pH (6).

Alpha-1 antitrypsin (AAT) is a 52-kDa serum protease inhibitor that is traditionally hard to purify. Conventional methods of manufacturing it follow ethanol precipitation procedures that produce only a 10% yield because the protein is inactivated by high ethanol concentrations (9). Because AAT is obtained from human blood plasma, the efficiency of extraction needed improvement. To that end, a VHH ligand was developed.

To assess the selected VHH ligand for functionality under the mild conditions required for this protein, AAT was purified from human plasma using a VHH affinity resin specific for it. AAT was successfully eluted from the resulting affinity resin with a neutral buffer (2 M MgCl₂, pH 7.4). To check the comprehensiveness of that neutral elution, the mild elution was then followed with an acidic elution (PBS, pH 2.0). No significant elution of product was observed in the acidic fraction, indicating that elution had been complete with 2 M MgCl₂. Fractions were analyzed using Coomassie-stained sodium-dodecyl-sulfate polyacrylamide gel electrophoresis (SDS-PAGE). The results in Figure 3 confirm that the affinity resin achieved a high-yield, high-purity purification of AAT from human plasma in a single step at neutral pH.

Lowering Cost and Improving Quality

A key benefit to using affinity chromatography for purifying proteins is the simplicity of the process, which can often reduce a complex multistep procedure to a single step that results in a high-purity, high-quality product. Step reduction not only reduces production time and associated costs, but decreasing the length of time a protein product is “in process” limits potential degradation or damage, thereby ensuring better final product. Most important, each step of a downstream process reduces final product yield, so it’s highly desirable to keep steps to a minimum.

Adenoassociated virus (AAV) is traditionally hard to purify. Over the past decade, it has received increased interest as a potential gene therapy vector. Current, conventional purification processes require multiple purification steps including density-gradient purification, several chromatography steps (ion-exchange, hydrophobic-interaction, heparin, and A-20 affinity), and an associated lengthy cleanroom residence time. With such a multistep process, the typical AAV yield is low. The process also suffers from poor scalability, limiting the commercial feasibility of this promising vector.

To address that challenge, a VHH ligand against AAV was identified and developed. Screening delivered a 14-kDa VHH fragment that was shown to bind AAV subclasses 1, 2, 3, and 5 and provide an eluate with high purity (7). Process steps were reduced from the typical four or five to just one or two steps, with a related reduction in cleanroom time for the recovery process from weeks to days. In this case, step reduction improved product recovery from 30% to >60%. Figure 4 shows the one-step purification of AAV serotype 1 using AVB Sepharose High Performance anti-AAV VHH ligands from GE Healthcare. The VHH ligand produced similar product purity to that of the reference standard.

Contaminant Removal

A highly desirable feature for all purification steps is the ability to remove contaminants such as host-cell proteins (HCPs) and DNA or virus particles. HCP and DNA clearance are good indications of the selectivity of purification media. In an analysis of protein A alternative chromatographic media for antibody purification, VHH ligands were found to be the only alternative that provided comparable HCP removal (10).

Viral clearance studies of the FVIII VHH ligand were performed by McCue, et al. (6). In their studies, clarified cell culture samples were spiked with one of two model viruses: xenotropic murine leukemia virus (X-MLV) to evaluate the adsorbent’s ability to remove retroviruses or retrovirus-like particles; and mouse minute virus (MMV) to evaluate the ligand’s ability to remove a small nonenveloped virus. Column elution pools were evaluated for virus presence, and polymerase chain reaction (PCR) was used to quantify the amount of virus in both the load and elution fractions. As Table 1 shows, the FVIII adsorbent provided significant clearance of both model viruses (6).

The ability of VHH ligands to remove HCP, DNA, and adventitious or endogenous viruses is expected to be good because of the ligand’s complexity. The more complex a ligand is, generally the more selective it is.

A Platform Technology

VHH ligands provide a highly flexible platform technology for purification of almost any target protein with high specificity and selectivity and good contaminant clearance under mild elution conditions. A number of VHH affinity adsorbents are commercially available already and being used for production of CTMs. Many more are being developed for future research and bioprocess use in the future (11).

Use of VHH media as a platform could shorten development times for purification of a given molecule by providing a plug-and-play solution for downstream processing. The simplicity and high specificity of the process reduces overall complexity in multiproduct facilities and ensures a minimum number of process steps. That increases yield while retaining excellent purity and quality of the end product. VHH ligands offer a clear advantage for multiproduct facilities and groups working with typically hard-to-purify molecules.

REFERENCES

1.) Liu, J. 2009. Comparison of Camelid Antibody Ligand to Protein A for Monoclonal Antibody Purification. BioPharm Int.:35-43.

2.) Detmers, F, P Hermans, and M. ten Haaft. 2007.Affinity Chromatography Based Upon Unique Single Chain AntibodiesThe Peak:7-14.

3.) Beyer, T. 2009. Serum-Free Production and Purification of Chimeric IgA Antibodies. J. Immunol. Meth. 346:26-27.

4.) Klooster, R. 2007. Improved Anti-IgG and HSA Affinity Ligands: Clinical Application of VHH Antibody Technology. J. Immunol. Meth. 324:1-12.

5.) Kuroiwa, Y. 2009. Antigen-Specific Human Polyclonal Antibodies from Hyperimmunized Cattle. Nature Biotechnol. 27:173-181.

6.) McCue, JT, K Selvitelli, and J. Walker. 2009. Application of a Novel Affinity Adsorbent for the Capture and Purification of Factor VIII Compound. J. Chromatog. A 1216:7824-7830.

7.) Hellström, M. 2008. Cellular Tropism and Transduction Properties of Seven Adeno-Associated Viral Vector Serotypes in Adult Retina After Intravitreal Injection. Gene Ther. 16:521-532.

8.) Smith, RH, JR Levy, and RM. Kotin. 2009. A Simplified Baculovirus-AAV Expression Vector System Coupled with One-Step Affinity Purification Yields High-Titer rAAV Stocks from Insect Cells. Molecular Therapy 17:1888-1896.

9.) Kumpalume, P. 2007. New Process for the Manufacture of α-1 Antitrypsin. J. Chromatog. A 1148:31-37.

10.) Low, D, R O’Leary, and NS. Pujar. 2007. Future of Antibody Purification. J. Chromatog. B 848:48-63.

11.) BAC’s Pipeline.