Recombinant Antibody Technology for the Production of Antibodies Without the Use of Animals
By Michelle M. Echko and Samantha K. Dozier, Ph.D. People for the Ethical Treatment of Animals (PETA)
Monoclonal antibodies (mAbs) are ubiquitous in biomedical research and medicine. They are used to fight, diagnose and research disease and to develop and test new drugs. While the mouse ascites method of mAb production is widely discouraged due to the substantial pain and distress involved and the equivocally named in vitro method has become standard, both methods raise serious animal welfare concerns. Fortunately, an alternative to animal-based mAbs exists. Synthetic antibodies called recombinant antibodies (rAbs) can be created using antibody genes made in a laboratory or taken from human cells, completely eliminating animals from the antibody-production process. rAbs can be used in all applications in which traditional mAbs are used and have inherent advantages over their animal-derived counterparts as well. The following review will discuss the production of rAbs, their advantages, and some impediments to their use.
Antibodies are proteins produced by immune cells in response to an invasion of foreign materials known as antigens. Antibodies protect us from infection by binding to foreign antigens and marking them for destruction or by binding to areas necessary for the antigens to function and directly interfering with their ability to cause harm.
Antibodies that come from a single immune cell are called monoclonal antibodies (mAbs). mAbs have become ubiquitous tools in biomedical science and medicine – they have been used to research, diagnose and combat diseases . Because mAbs have the ability to bind to cell-specific antigens and either target those cells for destruction or neutralize any deleterious effects they may have on the body, mAbs have emerged as effective therapeutic treatments for cancer, various auto-immune disorders, and other diseases. The Food and Drug Administration (FDA) has approved 22 monoclonal antibodies for therapeutic use and hundreds more are in various stages of clinical trials.
In the area of toxicological research, mAbs are frequently used as capture reagents to detect and measure protein and drug levels in biological fluids and to register changes in cellular proteins after exposure to a chemical agent has occurred. Antibody arrays are also becoming increasing popular platforms for the large scale analysis of proteins whose expression is modified by the administration of drugs.1
Monoclonal antibodies are so important to scientific research that several comprehensive initiatives for the generation and validation of mAbs have been undertaken worldwide. These initiatives include the Human Antibody Initiative, the European ProteomeBinders consortium, the German Antibody Factory, and the United States National Cancer Institute's (NCI) Clinical Proteomic Technologies for Cancer. The collective mission of these initiatives is to facilitate the characterization of all proteins produced in the human body by creating the antibody tools necessary to study them.
It is estimated that there are 20,000-25,000 protein-coding genes in the human genome2 and each such gene has the capability of making multiple different versions of a single protein. Creating antibodies for every variation of human protein is a huge undertaking that would entail the use of hundreds of thousands—if not millions—of animals with traditional mAbs technology.
First developed in 1975 by Kohler and Milstein,3 the conventional method for isolating and immortalizing individual immune cells that produce mAbs involves immunizing a mouse with an antigen of interest and then fusing the antibody-producing spleen cells of the mouse with immortal cancer cells. The fusion creates hybridoma cells that secrete antibodies into fluid that surrounds them as they grow. The hybridomas are either injected into the abdomen of a second mouse (the ascites method) or cultured in flasks or bioreactors (the in vitro methods) to produce large amounts of mAb-containing fluid.
It is well established that the ascites method of mAb production causes discomfort, distress, and pain to the animals involved, as the governments of Australia4, Germany, Switzerland, the Netherlands, and the United Kingdom5 have effectively banned it in favor of in vitro methods. In the United States, the National Institutes of Health and the U.S. Department of Agriculture both endorse in vitro methods as the default procedure for producing mAbs.6 Though laws and guidelines limiting the ascites method of mAb production in favor of in vitro methods are important and mark significant progress in the protection of animals in laboratories, it is important to note that in vitro methods involving hybridomas still require the immunization and subsequent euthanasia of animals, which presents animal welfare concerns.
These in vitro methods are also laborious, slow, and produce antibodies that often cause immune reactions in patients, limiting their use as clinical therapies . Mouse and other animal-derived antibodies must be altered, or "humanized", before they can be administered to humans. Humanization of antibodies is an imperfect process that is fraught with uncertainty. There is no guarantee that a humanized antibody will perform the same as its animal-derived predecessor and immune reactions varying in degree from skin rashes and hypersensitivity to reactions that are more sever may still occur.7,8,9
Fortunately, an alternative to the animal-based methods of monoclonal antibody production exists that can completely replace the use of animals. Synthetic antibodies called recombinant antibodies (rAbs) can be created using antibody genes made in a laboratory or taken from human cells, eliminating hybridomas and animals from the antibody-production process. Recombinant antibodies can be used in all applications in which traditional mAbs are used and have advantages of their own.
Production of Recombinant Antibodies
The advent of recombinant DNA technology10,11,12 laid the foundation for recombinant antibody production by enabling the combination of genetic material from two or more sources. In 1990 John McCafferty and colleagues demonstrated that antibody fragments could be displayed on viruses that infect bacteria, called bacteriophages or phages, by introducing antibody DNA into phage genomes via vectors.13 Researchers then created libraries of antibody genes to display on phages and developed methods to successfully isolate individual antibodies from the large phage-displayed libraries. This method has become known as antibody phage display. Of the 22 FDA approved therapeutic monoclonal antibodies currently on the market, one was created through non-animal antibody phage display technology. Of the previously mentioned antibody initiatives, both the German Antibody Factory and the European ProteomeBinders consortium create non-animal recombinant antibodies using antibody phage display. Since 1990, researchers have expanded antibody display technology to microorganisms other than phage. Koder and Wittrup, for example, documented the display of antibody fragments on Saccharomyces cerevisiae yeast in 1997.14
The production of non-animal recombinant antibodies can be broken down into five steps: (1) creation of an antibody gene library; (2) display of the library on phage coats or cell surfaces (3) isolation of antibodies against an antigen of interest; (4) modification of the isolated antibodies and (5) scaled up production of selected antibodies in a cell culture expression system.
Antibody Gene Libraries: Antibodies are Y-shaped molecules made up of heavy and light chains that each have a variable and a constant region. Schematic representations of antibody structure can be found here. The variable regions of both the heavy and light chains are responsible for antigen binding. The constant region of the heavy chain determines the class, or isotype, to which the antibody belongs. There are five different antibody isotypes in mammals (IgM, IgD, IgG, IgA, and IgE).Antibodies of the same isotype have the same constant regions.
An antibody gene library is a collection of microorganisms that have been transformed with the genes for the variable regions of different antibodies. The variable region genes can either be synthesized in vitro or amplified from the genetic material in human antibody-producing B cells. Variable region genes are used instead of genes for whole antibody molecules because fragments of antibodies are more easily assembled in microorganisms than whole antibody molecules and the variable regions of an antibody are the most important fragments in terms of function.
Each variable region gene is spliced into a vector (vehicles that transfer foreign genetic material into another cell), and the vector is inserted into a microorganism.
Library Display: Antibody library vectors contain genetic instructions to produce a protein found on the surface of viral particles or cellular membranes. Variable region genes are spliced into these instructions so that a fusion protein, composed of a fully functional antibody fragment and the surface protein, is created when the vector is inserted in a microorganism.
In yeast display platforms, antibody variable region genes are fused to the yeast Aga2p gene. The Aga2p protein naturally forms a linkage with a protein, Aga1p, that is anchored in the yeast cell wall. After the antibody fragment- Aga2p protein is expressed, it moves to the surface to link with Aga1p and the antibody fragment is displayed on the yeast cell surface as a result of the linkage.
In phage display platforms, antibody variable region genes are frequently fused to the pIII phage coat gene. The antibody library vectors are inserted into bacteria and the bacteria are infected with modified phage. The vectors and phages work together to disrupt the normal activity of the bacteria so that the bacterial cells begin producing new daughter phages that contain antibody genes and display functional antibody fragments on their surfaces. The antibody fragments are connected to the pIII phage coat proteins.
Progress is being made in the display of whole antibody molecules. Georgiou and colleagues have developed an E. Coli based display platform named E-Clonal that displays full-length IgG molecules, rather than fragments, on the bacterial inner membrane15 and several biotechnology companies that cater to the pharmaceutical industry advertise full-length antibody display platforms.16,17,18
Antibody Isolation: Once the rAbs are displayed, paramagnetic beads, fluorescence-activated cell sorting (FACS),19 and/or Enzyme-Linked Immunosorbent Assays (ELISAs)20 can be used to isolate individual antibodies that bind to a specific antigen target. The antigen of interest is incubated with the antibody-displaying microorganisms and the rAbs that do not stick to the antigen are discarded. Bound rAbs are removed from the antigen and screened for desirable characteristics.
Modification: Promising antibodies are grown in greater quantities and put through the selection process again to enrich for the highest performing candidates. If the affinities of the lead candidates are not strong enough, the antibodies can be "matured" through random or rational mutagenesis methods. Affinity maturation of an antibody is a process that takes place naturally in the body. Researchers use an analogous in vitro process to produce recombinant antibodies. Molecular biology techniques such as site-directed mutagenesis error prone PCR, DNA shuffling, or mutator strains of bacteria can be used to mutate selected residues of a given antibody fragment creating a whole new library that can be tested for increased function.21
Antibody Expression: Once a desirable antibody is chosen, the genes for the antibody are transferred via expression vectors into an expression system—bacteria, yeast, or mammalian cell lines specially designed for the expression of foreign proteins. The choice of vector and expression system depends on the type of antibody that is to be produced.
Antibodies from antibodies libraries are usually monovalent fragments that contain only one antigen binding site and no constant region. Natural antibodies have a copy of the antigen binding site on each of the Y arms (2 binding sites in total) and a constant region at the base of the Y. Expression vectors are available that will connect two or more fragments together to create divalent and multivalent antibody molecules. Expression vectors can also add a constant region to fragments to recapitulate the full length antibody structure.
Full length antibody expression is most frequently done in yeast and mammalian expression systems. Lonza's CHOK1SV, Percivia's PERC.6 and HEK293 are some of the mammalian cell lines currently in use for the production of whole recombinant antibodies. All of these lines have been reported to yield rAbs in the grams per liter medium-range.22,23 Bacterial expression systems are quick and inexpensive but have mostly been used to express fragments rather than full length antibodies.
Advantages of Recombinant Antibody Use
Non-animal technology: Recombinant antibodies derived from synthetic or human antibody libraries are an entirely non-animal technology. This alleviates animal welfare concerns associated with traditional monoclonal antibody production (animal immunizations, euthanasia and ascites hybridoma expansion).
Speed: Once an antibody library is established, a researcher skilled in the art of recombinant antibody production can furnish an antigen-specific antibody suitable for research purposes in as little as 8 weeks.24 This is significantly shorter than the 4 or more months required for hybridoma technology. Speed is a boon to facilities that produce numerous antibodies – specifically antibody vendors, university core laboratories, and laboratories engaged in proteomic research.
Control: Recombinant antibody production gives researchers control over the state of the antigen to which they are making antibodies against. With traditional mAb technology researchers lose control after injecting the antigen into an animal. Inside the animal an antigen could be processed into something different or cut into pieces and antibodies may be made against these altered versions of the antigen instead. Animal-derived antibodies must be tested extensively after they are created in the hope that an antibody against the intended antigen has been made. With recombinant antibody production, researchers can guide the production process by adjusting experimental conditions to favor the isolation of antibodies against specific antigens or antigen characteristics. For example, Wells and colleagues were able to derive antibodies through phage display technology that distinguish between the activated form of the caspase-1 enzyme and the inactivated form.25 To do this, caspase-1 molecules were chemically modified to maintain an activated state. The activated caspase-1 molecules were mixed with a phage display library, bound antibodies were removed, and these antibodies were tested against chemically inactivated caspase-1 molecules. Only the antibodies that bound exclusively to the activated caspase-1 molecules were kept; antibodies that bound to both the activated and inactivated molecules were discarded. Had animals been used to create antibodies in this experiment, the caspase-1 antigens might have lost their activated and inactivated conditions and antibodies capable of distinguishing between them would not have been created.
Isotype Conversion: Once a desirable antibody fragment is found it can be easily converted into any antibody isotype (e.g. IgA, IgM IgG etc.) from any species by adding the appropriate constant domain.26
Applications and Other Advantages: Recombinant antibodies obtained from antibody gene libraries can be used in all applications in which traditional mAbs are used (e.g., western blotting, immunohistochemistry, fluorescence-activated cell sorting (FACS), and immunofluorescence),27 and they have several more advantages over their traditional animal-based counterparts as well.
Recombinant antibodies from antibody gene libraries require less purified antigen to produce28 and eliminate constraints on the types of antigens that can be used. Antibodies to highly toxic or non-immunogenic antigens can be created using library methods, unlike animal immunization technologies.29 Furthermore, the affinity of antibodies derived from libraries can be increased to levels unobtainable by an animal's natural immune system.
Recombinant antibodies from antibody gene libraries are also amenable to high-throughput production30,—a fact that makes them attractive to the field of proteomics. Because they are derived from human or synthetic genes, recombinant antibodies do not trigger the intense immunogenic reactions in patients that animal-derived antibodies do31, making them ideal for clinical use. As one researcher from Tufts University pointed out, "[T]he number of applications in which rAb technology has advantages over mAb technology is becoming too numerous, and the power of these applications too great for many labs to resist.32
Impediments to Recombinant Antibody Use
Intellectual property: Recombinant antibodies represent the fastest growing segment of today's pharmaceutical industry.33 As a consequence, commercial interests have restricted access to improved methods of generating rAbs to protect their intellectual property. Many of the technologies that enable generation of recombinant antibodies are only accessible through business partnerships or potentially expensive licensing agreements.34
Initial Investment: Recombinant antibody technology requires a substantial investment of time and energy in order to implement; laboratories needing only a few antibodies may be unwilling to make the investment to learn and implement this new technique. Researchers must design and create the equivalent of an in vitro human immune system –the antibody gene library- from which they will pick out antibodies that bind to their targets of interest. To be effective, an antibody library must be equal to or surpass the diversity of antibodies found in a natural immune system and this diversity must be obtained without sacrificing functionality. Synthetic libraries created through methods of random mutagenesis can contain large numbers of nonfunctional antibodies that are unable to bind a target. Libraries of human antibody genes taken from B cells increase the probability of obtaining functional genes but large numbers of human volunteers can be required to populate the library. The inconvenience of the latter technique has in part been offset by cataloging genetic information from volunteers in publically accessible bioinformatic databases, allowing easy retrieval of the information for future library generations. In either event, the design of a large, useful antibody library requires detailed structural and functional knowledge of proteins in general and of antibody molecules in particular.
Technical Expertise and Robotics: The library design step can be bypassed by acquiring aliquots of a proven, pre-made library. However, sorting through the immense libraries may also present an obstacle to embracing this technology.
It is difficult to evaluate the very large numbers of candidate antibodies that come with an antibody library without robotics such as colony pickers, automated ELISA machines, and FACS machines, all of which are costly pieces of equipment.
Selection and screening of antibodies from antibody libraries requires technical expertise in molecular biology especially if phage-displayed libraries are used. Bacterial transformation, phage titration, phage amplification and DNA mutagenesis are some of the skills necessary for isolating quality antibodies from phage-displayed libraries. Though these skills are standard in labs conducting molecular biology work, labs based more in cell biology and immunology may find them difficult to incorporate.
Phages themselves are inherently sticky and may bind antigens regardless of whether or not they bear an antibody corresponding to the antigen. This can lead to false positives if measures are not taken to minimize non-specific binding. For these reasons, it is important to choose a robust strategy for selecting antibodies from a display library and use multiple rounds of selection to separate binding from non-binding clones. Blocking agents should be used in the reaction mixtures to curb non-specific binding.
Maturation and Structure Conversion: Upon repeated exposure to an antigen, an animal will create progressively stronger antibodies against it. Antibodies generated during the primary response to an antigen are mutated and a secondary response is mounted. This process, called affinity maturation, may repeat itself several times until strong antibodies that can neutralize an antigen or eliminate it from the body are generated. Traditional monoclonal antibody technology takes advantage of an animal's natural affinity maturation process by immunizing an animal several times with the same antigen over a period of weeks. Recombinant antibody technology requires that researchers carry out the affinity maturation process themselves with advanced molecular biology techniques.
Affinity matured antibody fragments selected from an antibody library may also need to be converted to divalent or full-length antibodies to increase their utility in specific applications.
The use of rAbs is scientifically validated and has inherent benefits not available in animal-based antibody production methods, which present a host of methodological and ethical concerns. Recombinant antibody technology is sufficiently advanced to allow for its evaluation and implementation in laboratories.
The letter and spirit of animal welfare laws governing animal experimentation in the U.S., E.U. and elsewhere stress the importance of seeking, considering and implementing modern alternatives to the use of animals. Recombinant antibodies from synthetic or human antibody libraries are a legitimate and, in many applications, a superior alternative to animal-based methods of monoclonal antibody production and they are not being used as frequently as they should. While it is true that there are obstacles to the widespread use of non-animal rAbs, these obstacles are not insurmountable. If replacing animals in experiments is a priority for the research community, it is not sufficient to abandon only the ascites method of hybridoma expansion (though this action is necessary and laudable). Steps must also be taken to embrace challenging but entirely non-animal advances like those offered by recombinant antibody technology.
The Appendix lists recombinant antibody resources that may aid researchers in their investigation of this technique. Included in this list are academic core facilities that maintain antibody libraries and that are willing to generate custom recombinant antibodies for outside researchers.
©2010 Michelle Echko & Samantha Dozier