Antibodies play a central role in many of the key techniques that drive biological research these days. So perhaps it’s not surprising they also have become a focal point in the drive for greater reproducibility in science.
By one estimate, “poorly characterized and ill-defined antibodies” account for $350 million in wasted research in the United States each year, and $800 million globally [1]. There’s any number of potential reasons for that: Antibodies may not be as specific as intended, or they may not function properly under certain experimental conditions. There’s also no guarantee that one vendor’s antibody to a given antigen will perform the same as that of a competitor, nor even that different batches of the same antibody will work identically.
Researchers and opinion leaders have considered a number of potential solutions to these problems. One, outlined in a 2015 comment in Nature, is to transition from traditional monoclonal or polyclonal antibodies to recombinant antibodies [1].
In a 2015 Survey conducted by Biocompare (n=884 respondents), researchers were asked what advantages they found in using recombinant antibodies vs. other, non-recombinantly generated antibodies. The top three responses were: 1) improved target specificity, 2) greater sensitivity and 3) lot-to-lot consistency.
According to Anthony Couvillon, scientific project manager at Cell Signaling Technology (CST), a recombinant antibody is simply an antibody (or, more broadly, a “binder”) whose coding sequences have been cloned and expressed in a mammalian cell line. Thus a recombinant antibody is conceptually equivalent to any other recombinant protein.
Because a recombinant antibody is a unique molecular species, it is most similar to a standard monoclonal antibody.
But with traditional monoclonals, which are produced by fusing antibody-producing B cells with immortalized melanoma cells, the antibody is only as good as the B cell that produced it; there’s no opportunity to tweak its affinity, for instance.
“In recombinant antibodies, you are expressing them in specialized cells, and that allows you to tinker with the [coding] chain,” Couvillon says.
How are they made?
Fundamentally, recombinant antibodies are made in one of two ways. Some are generated, as traditional antibodies are, in vivo, by immunizing animals. However, instead of simplify purifying sera or making hybridomas, recombinant-antibody developers screen the resulting B cells to identify high-quality binders, and then clone the associated immunoglobulin heavy and light chain genes into mammalian expression vectors and transfect those into non-B cell lines (such as 293 or CHO) to ramp up production.
The alternative approach is to produce the molecules in vitro. But in this case, the binders are not typically full-fledged IgG molecules but rather smaller designs, such as scFv or Fab fragments.
IgG is a large, complex structure comprising two copies each of two separate proteins linked by disulfide bridges and adorned with multiple post-translational modifications. That makes the molecule relatively difficult to manufacture and screen. An scFv (single-chain variable fragment) combines the two antigen-binding regions of the heavy and light chains in a single, far smaller molecule; Fab (antigen-binding) fragments are somewhat larger molecules, with the two chains encoded by separate genes.
Abcam’s RabMAb primary antibodies and Thermo Fisher Scientific’s ABfinity reagents are generated in vivo, as are many antibodies from CST. “About 46% of our total antibody portfolio is recombinant,” including nearly 90% of the company’s monoclonals, says CST’s Couvillon. Polyclonals, by definition, cannot be recombinant, Couvillon notes. However, CST does bundle multiple recombinant monoclonals in defined ratios to produce polyclonal-like preparations called MultiMabTM mixes, which are especially useful for detecting post-translational modifications.
For in vitro generation, researchers build randomized libraries in their protein architecture of choice, screen them against the targeted epitope in a so-called phage display or similar assay and then collect, propagate and retest the resulting hits to isolate the highest-affinity binders. Among other things, this strategy is considerably faster than traditional in vivo antibody production, taking weeks rather than months, according to Amrik Basran, chief scientific officer at Avacta Life Sciences. It also allows researchers to develop antibodies to molecules that might be harmful to a host (such as toxins) or to novel sites on a target antigen that may be difficult to generate an antibody to. In addition, this approach lets researchers use negative-selection steps to eliminate binders targeting closely related sequences, thereby preventing off-target binding events.
Once created, these in vitro-generated molecules can be used as is, or (depending on the design) they can be cloned back into an immunoglobulin backbone to take advantage of the many tools available for researchers using IgG, such as protein A-conjugated beads and secondary antibodies, which do not necessarily exist for alternative architectures. “There’s a whole ecosystem of reagents around [immunoglobulins] that are not necessarily available for other designs,” says Gary Dillon, head of technology and research and development at Abcam.
However they originate, recombinant antibodies can be optimized post-screening to boost affinity for their target antigen.
But at Bio-Rad Laboratories, few clients find it necessary to take the company up on that offer, says Andrea Murphy, Custom Antibody Specialist at Bio-Rad's Antibodies division. Its custom-made recombinant antibodies average “low nanomolar to high picomolar” affinities. “With affinity maturation, we can reach low picomolar affinities.” The company also sells recombinant antibodies off-the-shelf, says Laura Moriarty, Drug Discovery and Development Marketing Manager at Bio-Rad, and while the current catalog of such molecules is relatively small, "they are often top sellers, such as our range of anti-biotherapeutic antibodies," she says.
Abcam’s recent acquisition, AxioMx, builds recombinant antibodies on a single-chain antibody backbone called scFv (single-chain variable fragment). Others rely on different molecular structures: ankyrin repeats, in the case of Molecular Partners; and cystatins, in the case of Avacta Life Sciences.
Cystatin, Basran explains, is a human protein whose normal job is binding and regulating the proteinase cathepsin. Into that binding interface, Avacta engineers two nine-residue loops, each of which is randomized (note: the whole loop is random, not just one position). The resulting molecules no longer binds cathepsin, but some, at least, will bind to the target of interest, Basran says: The number of possible combinations is “vast.”
Furthermore, full-fledged antibodies are 150-kDa heterotetramers, but Avacta’s Affimers, Basran says, are only about 14- to 15-kDa single chains—easy to produce in bacteria and devoid of disulfide bonds. “So they are easier to manufacture.”
Miltenyi Biotec offers its REAfinity™ recombinantly engineered antibodies (over 5000 options) that are developed and validated for flow cytometry analysis. These antibodies are generated in standardized culture conditions in mammalian cells, but are still highly pure and free from hybridoma derived contaminating immunoglobulin impurities. REAfinity Antibodies include the same engineered human IgG1 constant region allowing the use of one isotype control for all REAfinity Antibodies and lack binding to Fcy recptors which results in a higher signal specificity during cell analysis. The portfolio of antibodies includes markers for chemokine receptors, cell signaling proteins and common immune antigens such as CD3, CD4 and CD56.
Some developers, including ChromoTek, Capralogics and Creative Biolabs, design recombinant antibodies on a camelid antibody scaffold. According to Klaus Herick, head of global commercial operations for ChromoTek, camelids (which include llamas, camels and alpacas) produce simple antibody variants called heavy-chain antibodies, which lack the light chain found in mammalian antibodies.
ChromoTek takes the single variable domain of these antibodies, called VHH, and couples it to beads, fluorophores or fluorescent proteins to create small binding reagents called Nano-Traps, Nano-Boosters and Chromobodies®, respectively. Among other things, these reagents may offer an advantage in super-resolution microscopy: Their diminutive size (just 3 or 4 nm in diameter) means the fluorophore is positioned closer to the targeted structure than is possible with a full-sized antibody molecule, Herick explains.
Alternative options
Other companies also sell recombinant binders, including Absolute Antibody, MilliporeSigma and R&D Systems. In some cases, researchers can specifically search for recombinant molecules; in other cases, they are simply mixed into the catalog like any other antibody.
So should researchers care how their antibodies are produced? Yes and no. On one hand, if a reagent works, it works. But polyclonals are notoriously variable from lot to lot, and monoclonals may be so specific as to be inflexible across applications. Furthermore, hybridomas, from which monoclonals are made, require considerable care and maintenance, and they can degrade over time, for instance via “genetic drift.”
As a result, researchers with a strong interest in ensuring a specific antibody will be available for the foreseeable future are well advised to look for recombinants, says Matt Baker, director of research and development and business development for research-use-only antibodies and immunoassays at Thermo Fisher Scientific. “The great thing about a recombinant antibody is you have the DNA, and the DNA is stable forever.”
There are other advantages, too. Given the genetic sequence of an antibody, researchers can tweak it, for instance to boost antigen affinity, add conjugation sites or introduce non-natural amino acids. They also can experiment with antibody structure, for instance to build bispecific or other high-valency binders and thus target multiple pathways simultaneously. (Or, more likely, they can work with a service provider to create such antibodies on their behalf; recombinant antibody creation and production is “quite extensive and involved,” Baker notes.)
Many of these alterations are, of course, possible with traditional antibodies. But if nothing else, it’s a lot of work. Besides, going recombinant represents “the ultimate banking method,” says Dillon. Maintaining hybridomas requires serious effort, he says. “Why do that when you can sequence them and then bank your antibody in silico? If your facility burns down, you’re not worried. You’ve got the sequence.”
Reference
[1] Bradbury, A, Plückthun, A, “Reproducibility: Standardize antibodies used in research,” Nature, 518:27-9, February 5, 2015. [PMID: 25652980]