Hybridoma VS Display : The Fight Of The Century

Once you have (finally!) identified your antigen and want to develop a new monoclonal antibody against it, the same questions always arise. We all have faced the same anxiety asking ourselves « should I use display or traditional hybridoma procedure to generate my new antibody ? » « Is a monoclonal antibody generated through the latter technique will show a greater affinity toward the target ? » « And regarding a long-term view on my project, what is the most secure approach for next years ? » .

So let’s tackle these questions one after the other while making a little refresh about monoclonal antibody generation procedures.

Hybridoma generation was first developed in 1975 by George Köhler and César Milstein. It relies on the fusion of mouse immunized B spleen cells with myeloma cells. The immortal B cells producing the antibody of interest are then selected and the best clones are screened in order to obtain monoclonal antibodies with the best antigen affinity.

Only a few people know it, but it is at the same period that Hervé Bazin created the rat myeloma cell line IR983, allowing the first generation of rat monoclonal antibodies in Louvain-la-neuve, Belgium, and so given birth to SYnAbs company.

In general, the development of hybridoma coming from new species is limited due to the lack of stable fusion partner. Fusing B cell from one specie with myeloma cell line from another one will give birth to hetero-hybridoma clones which will be gradually lost during the clonal selection step due to their genetic instability.

Hybridoma development remain one of the more traditional and robust methods used to generate monoclonal antibodies, since :

• it allows you to produce monoclonal antibodies in a cost-effective, weakly labor intensive and poorly technically demanding way,

• it is important to note that the mammalian origin of the cells integrates in vivo post translational modifications, which decreases the risk of potential aggregation,

• it allows the generation of highly specific monoclonal antibodies to a single epitope of the target antigen,

• natural affinity maturation occurs in animal. Complex in-vivo recombination processes ensure the optimal combination of variable and constant domains, resulting in antibodies with very high affinity features.

However, there are some significant drawbacks to using hybridoma technique :

• the hybridoma development process is long. Counting from immunization, it takes on average between 3 and 4 months to obtain a hybridoma clone. 

• the animal origin of the antibodies implies further humanization for a therapeutic purpose. Hybridoma derived antibodies cannot be improved until they are first converted into recombinant antibodies or raised through transgenic animals,  which induces an extra cost in any case.

• there’s no control over the epitopes to which antibodies are formed.

• hybridoma technique requires expensive robotics (or long hours in the lab, your choice ;-) to screen thousands of antibody candidates. The screening step is known to be very laborious and delicate. 

Other limitations are directly linked to the nature of the antigen :

• sensitive  antigens  (e.g.  difficult to express proteins, intracellular compounds, conformational epitopes…) could be impossible to target,

• toxic  antigens  may  kill  the host  animal  before  antibodies are even produced.

To overcome these limitations, SYnAbs has developed a unique DNA immunization technique allowing the expression of the native antigen and the screening on cells via Flow Cytometry FACS machine.

Sometimes, proteinic antigens are highly conserved between species and may not elicit an immune response.

Finally, hybridoma are known to undergo genetic drift leading to batch-to-batch variability.

With the spread of low-cost sequencing techniques, it is now possible to overcome this instability by sequencing the hybridoma and saving genes of interest into a plasmid vector. This is typically what RD-Biotech offers, passing from hybridoma to other expression systems (e.g. Chinese hamster ovary cells, CHO).

But with hybridoma technology, how can we :

• simultaneously generate multiple binders against different antigens with a poor screening throughput ?
• explore human repertoires or repertoires from species that have not been widely used for hybridoma production ?
• perform further genetic engineering of the binding sites, i.e. affinity maturation, specificity optimization, mutagenesis scanning… ?

Facing these limitations, G.P. Smith introduced in 1985 the concept of displaying exogenous proteins on the surface of filamentous M13 bacteriophage, showing the potentials of building phage libraries displaying large repertoires of different proteins.

A phage displaying a specific antibody on its surface can be isolated for its binding property to a target ligand starting from a collection of billions of phages displaying different antibodies. Since the phage displayed protein gene is present in the phage genome, the selection of a virus allows the concomitant recovery of the corresponding antibody gene. Once isolated genetic details are easily identified by DNA sequencing and the sequence could be used for subsequent applications.

First, a library containing the antibody DNA sequences is created. Antibody diversity is restricted to the variable regions (VH and VL) and these gene fragments are inserted into a specific vector with the sequence encoding the phage protein pIII. Once assembled, the phage particle will expose the functional antibody fragment fused to the amino terminus of the minor coat protein III.

In the creation of an antibody library several different choices can be made:

• which form of antibody fragment to use (single-chain fragment variables scFv or Fragment antigen-binding Fab)
• the source of V regions repertoire, recovered by RT-PCR amplification starting from lymphocytes which may (hyperimmune library) or may not (naive library) have undergone antigen stimulation.

Once a library is created, the enrichment of antigen-specific phage antibodies is carried out by “phage panning”, using immobilized or coupled antigens (to magnetic beads).

Unlike hybridoma techniques:

• When a naive library is already available, this process can be very fast and generally lasts only a few weeks,

• When a library is set-up, it can be used multiple times and be panned again several years later,

• Phage display is commonly used for toxicology and antivenom research due to its ability to facilitate work with both toxic and non-immunogenic antigens,

• Phage display has a much higher throughput than hybridoma, because billions-diverse phage libraries can be panned.

But phage displays is not the perfect solution :

• A first limitation often observed, in comparison to hybridoma development, is the cost of the technique since phage display is more expensive and required a different set of operational skills.

• There is a chance that the binders may result in a moderate affinity to the targets when panning naive libraries. Therefore, to find good antibodies against diverse antigens, these libraries need to be very large. This makes the generation of naive natural libraries very laborious.

• Another potential limitation of antibody libraries is the VH-VL pairing. The natural pairings of heavy and light chains found in antibody-producing B cells are not retained as the combinatorial libraries are constructed, leading to a random pairing of VH and VL. This could potentially lead to the generation of antibody molecules that have suboptimal biophysical characteristics and lead to developability issues (aggregation, lack of solubility,…). 

• Another difference from hybridoma technology is that the output of display systems include scFvs or FAbs antibody fragments and they cannot recruit components of the immune system for cytotoxic effects through antibody-dependent cell-mediated cytotoxicity (ADCC) and complement dependent cytotoxicity (CDC).

• While the immune system generally does not create antibodies with monovalent affinities better than picomolar (10-12) range, affinities of recombinant antibodies can often be improved beyond this limit using in vitro affinity maturation. For this process, a known antibody sequence is diversified either completely, by random mutagenesis using a technique such as error-prone PCR, or the diversity is limited to certain CDRs, which are exchanged with CDR libraries or are randomly mutagenized. These derivative libraries are then used for additional rounds of selection. Affinity improvement of 5000-fold has been reported with this technique.  

The generation of antibodies against self-antigens is normally not possible with immune libraries, although knockout animals have been already successfully used for that purpose.

In antibody development, there’s no such thing as “one size fits for all”, display and hybridoma technologies are complementary approaches that increase the success rate of project. And this is why mAbexperts group has been created.

Both approaches have theoretical and practical limitations, although the in vivo approach has yielded most of the approved therapeutic mAbs so far.

In recent studies, researchers show that aliphatic contents of HCDR3 and HCDR2 were significantly higher in phage display derived antibodies (higher leucine and isoleucine content) compared to hybridoma.

Consequently:

• Phage display derived antibodies show higher cross-reactivity compared to hybridoma derived ones (confirmed via PSR binding assay, BVP ELISA). Because aliphatic residues lack H-bond/charge-charge interaction capabilities and definitive geometries (unlike aromatic residues) essential for complementarity, enrichment of them in the binding interface leads to poly-specificity.

• binding interface area is larger with phage derived antibodies and number of contacts are greater. For instance, adalimumab (trade name Humira®) interacts simultaneously with two monomers of TNFα trimer, where infliximab (trade name Remicade® issued from hybridoma technology) forms interaction with only one monomer of the natural trimer form.

• Aliphatic residues also contribute to aggregation due to their hydrophobic nature. Phage display derived antibodies might be more prone to aggregation.