What essential steps are necessary to accelerate recombinant antibody expression?

Factors influencing the choice of expression system and transformation/transfection strategy

The vast majority of antibodies currently in use for therapy and research are produced by different prokaryotic and eukaryotic recombinant systems.

Choosing the right system for each antibody depends on several factors including:

• Antibody format: full-length, bispecific, fragment, fusion protein, among others
• Intended use: therapy, diagnostics or research
• Required amounts
• Post-translational modifications
• Desired timeline for project completion, among others

Thus, therapeutic antibodies, which are typically full-length IgG molecules requiring complex post-translational modifications, large-scale and long-term production, are predominantly expressed by stable mammalian recombinant systems.

On the contrary, antibodies intended for research and diagnostics, can forgo complex glycosylation, and thus are easily produced in large quantities using simpler systems, such as bacterial and yeast expression systems.

Nevertheless, choosing the right system is only the first stage of this complex process.

In fact, many other factors play a key role in guaranteeing a successful antibody expression project. In many instances, a poor sequence design, choice of inadequate vector and suboptimal culture conditions, often lead to reduced antibody yields which, in turn, results in reduced final purity.

Vital steps in recombinant antibody expression

All antibody expression projects comprise different stages:

1. Choosing the adequate expression host
2. Designing and optimizing the expression vector
3. Defining the gene delivery strategy
4. Choosing between transient or stable transfection/transformation
5. Optimizing culture conditions
6. Scaling-up
7. Defining the purification strategy

At every stage, researchers need to weigh and choose the conditions that are better adapted to the ultimate goal of the project. Moreover, each decision conditions the subsequent steps of the process. For this reason, the most important decisions must be made at the very beginning of the project, namely, during steps 1 to 4.

Choice of expression host

Typical host expression systems for recombinant production include:

• Bacteria: Escherichia coli and Bacillus subtilis
• Yeast: Saccharomyces cerevisiae and Pichia Pastoris
• Insect cell lines:

• • Cell lines from the order Lepidoptera, moths and butterflies: TN-368, BTI-Tn5B1-4, Tni PRO, IPLB-Sf21AE, Sf900+, Bm-N
  • Cell lines from the order Diptera, flies: S2 and S2R+ both derived from Drosophila melanogaster
• Mammalian cell lines derived from: Chinese hamster ovary (CHO) cells, human embryonic kidney (HEK293), baby hamster kidney (BHK-21) and mouse myeloma (NS0)
• Plants: plant tissues or cells from different species including Nicotiana tabacum BY2, Nicotiana benthamiana or Daucus carota are recurrently used for recombinant expression

The choice of expression host ultimately depends on the structure of the antibody and its intended use. For instance, IgG-like antibodies intended for therapy require complex and human-like glycosylation. The structure and nature of these glycans are essential to guarantee the efficiency and safety of these biopharmaceuticals.

Hence, bacterial systems, which cannot produce any type of glycans, are generally avoided for the production of IgG-like biopharmaceuticals. Unlike bacteria, yeast systems can perform glycosylation, but the structure of these glycans is significantly different from those of human origin. Moreover, antibodies of yeast origin are known to possess a significantly low in vivo half-life. Thus, for this reason, yeast systems are also not recommended for the large-scale production of these biopharmaceuticals.

Nevertheless, bacterial and yeast systems are still widely useful for the production of antibody fragments that don’t require glycans (e.g. VHH, Fab, etc.) or IgG-like antibodies for research and analytical applications. The reason for this marked preference stems from the fact that these recombinant production systems are easily scaled-up and rely on the fast-transient transformation with selectable vectors.

But when it concerns IgG-like antibody expression, mammalian systems are still vastly favored. The reason for the marked preference relies on the fact that most mammalian systems (e.g. HEK293, CHO, etc.) can perform human-like glycosylation. Moreover, historically, mammalian cell lines such as CHO and HEK293 systems have been the object of many studies. Thus, on a genetic level, the methods for transfection and isolation of positively transfected cells are well established.

But, increasingly, scientists are exploring alternative systems for antibody expression, such as plant and insect cells. In recent years, much progress has been made on the large-scale antibody expression in plant cell lines or tissues. The main advantages of these systems are their low cost, high potential for scalability, and increased safety. Nevertheless, on an industrial level, there are few the number of antibodies already produced on these platforms. The most well-known example of large-scale therapeutic antibody expression in plants is ZMapp, a cocktail of three different monoclonal antibodies developed against infections with the Ebola virus.

Relatively less is known regarding antibody expression in insect systems. Although many biotherapeutics are currently produced in insect systems on a commercial level, these systems cannot perform human-like glycosylation. Hence, insect-derived antibodies may be immunogenic. For this reason, the applicability of these systems for IgG-like antibody expression is still quite limited.

Vector design and transformation/transfection strategy

The choice for a specific expression host ultimately influences the construction of the expression vector. Moreover, due to the inherent redundancy of the genetic code, organisms show great variability in their corresponding codon usage bias. Thus, to optimize expression, it is necessary to adapt the codon usage of every gene to the codon bias of each specific host. Finally, vector construction also depends on the specific mode of transformation/transfection that will be employed for a specific project.

Introducing foreign DNA into an expression host, and stabilizing the replication and expression of that new element is still considered one of the most challenging steps of modern recombinant technologies.

In bacterial and yeast systems, plasmids containing the gene of interest can be easily stabilized by including a simple selective element, and an origin of replication (ORI). The most commonly used selective elements are antibiotic or toxin resistance genes. The inclusion of these genes simplifies clonal selection, because, only successfully transformed cells are able to grow in the presence of the biocide. Moreover, the presence of an ORI in the expression vectors ensures the continuous replication of that element alongside the replication of the host’s genome.

On the contrary, in mammalian cells, in which plasmids do not usually occur, expression and replication are much harder to stabilize. Mostly because mammalian cells do not recognize widely characterized ORI genes and thus fail to replicate foreign plasmids alongside its genome.

Hence, in time, this causes the dilution or degradation of the expression vector lending to a progressive reduction of production yields and subsequent loss of the vector after only a couple of days (24-72 hours for RNA vectors, 48-96 hours for DNA vectors).

This progressive degradation of the new element is not as critical for mammalian cells and as it would be bacterial and yeast systems. The reason for this is that mammalian cells double at a much slower rate than bacteria and yeast. Thus, this apparent limitation can be played to our advantage by allowing the development of high performing transient systems able to achieve good levels of antibody expression in very short timeframes (3 to 7 weeks).

These systems are thus termed transient, as they require only the temporary incorporation of the accessory genetic elements. Nevertheless, as these systems ultimately lose the expression vector, it became necessary to develop alternative methodologies for the stable and long-term expression of antibodies at a clinical scale.

But, to stabilize expression levels in mammalian systems, it is necessary to achieve the irreversible integration of the new genetic element into the genome and perform the time-consuming selection of the antibody-expressing cell lines. Unfortunately, this need for genome integration and the inherent slow growth rates of mammalian systems, makes stable cell line development a very complex, labor-intensive and time-consuming process.

Because integration is a necessary step in stable cell line development, the vectors used for this approach need to be fundamentally different and more complex than the ones used for transient transfection or bacterial/yeast transformation. In fact, instead of an ORI, these vectors should contain the genetic machinery that allows the integration of the antibody-encoding genes into the genome. Interestingly, the selective elements of these integrative vectors also differ from the ones used for bacterial and yeast transformation.

Unlike bacteria and yeast vectors, which rely on biocides for selection, the most popular genes used for selection of stable selection exploit the intricate metabolic needs of the mammalian cell. These selective elements usually encode for:

• Dihydrofolate reductase (DHFR), involved in nucleotide metabolism
• Glutamine synthetase (GS), an enzyme responsible for regulating levels of ammonia in different tissues

In both cases, selection can be caused by subtracting the vital metabolite in the culture medium: hypoxanthine and thymidine for DHFR or glutamine for GS-based selection. Their absence prevents the growth of non-transformed cells and leads to the enrichment of positive cell lines.

Subsequently, positive single cells are transferred to a new cultivation medium which creates clonal populations able to constitutively express the antibody. From this stable pool, only the clones with the highest productivity and stability are chosen for subsequent optimization.

Stable cell line development can take from 6 to 9 months. Thus, these protocols are reserved for antibody expression at the clinical and industrial scale.

Concluding remarks

Antibody expression is a complex process involving multiple steps. Currently, the development of therapeutic antibodies starts by the short-term production of these molecules on transient mammalian systems, and ultimately requires the development of stable mammalian cell lines for clinical evaluation and commercialization.

At an industrial level, stable and transient expression in mammalian cells is still considered the gold standard for therapeutic antibody expression. The reason for the marked preference towards these systems results from their ability for performing proper protein folding and post-translational modifications. Two properties that greatly influence the safety and efficiency of therapeutic antibodies.

On the contrary, antibody expression for research and diagnostics is recurrently achieved in simpler systems, such as bacteria and yeast, in which the replication and stability of expression vectors can be easily achieved by including an adequate ORI and an antibiotic/toxin resistance gene.

1. Cérutti, M. and Golay, J. Lepidopteran cells, an alternative for the production of recombinant antibodies? MAbs. 2012; 4(3): 294–309. doi: 10.4161/mabs.19942
2. Clausen, H. et al. Glycosylation Engineering. 2017. In: Varki A, Cummings RD, Esko JD, et al., editors. Essentials of Glycobiology. 3rd edition. Cold Spring Harbor (NY): Cold Spring Harbor Laboratory Press; 2015-2017. Chapter 56. doi: 10.1101/glycobiology.3e.056
3. Geisler, C. and Jarvis, D. L. Adventitious viruses in insect cell lines used for recombinant protein expression. Protein Expr Purif. 2018; 144: 25–32. doi: 10.1016/j.pep.2017.11.002
4. Kim, T. K. and Eberwine, J. H. Mammalian cell transfection: the present and the future. Anal Bioanal Chem. 2010; 397(8): 3173–3178. doi: 10.1007/s00216-010-3821-6
5. Kingston, R. E. et al. Amplification Using CHO Cell Expression Vectors. Curr Protoc Mol Biol. 2002; Chapter 16: Unit 16.23. doi: 10.1002/0471142727.mb1623s60
6. Rademacher, T. et al. Plant cell packs: a scalable platform for recombinant protein production and metabolic engineering. Plant Biotechnol J. 2019; 17(8): 1560–1566. doi: 10.1111/pbi.13081