The idea that antibodies could serve as therapeutics emerged over 100 years ago. Nevertheless, antibody therapeutics have taken a long time to come of age. At first, antibodies were poorly understood, despite daring experimental therapies, such as the anticancer serotherapy introduced by Charles Richet and Jules Héricourt in 1895, and inspired theoretical models, such as the side-chain hypothesis offered by Paul Ehrlich in 1895.

Fortunately, researchers persisted. After decades of effort, they succeeded in developing versatile antibody therapeutic platforms. A significant advance was recognized in 1984, when the Nobel Prize in Physiology or Medicine was awarded to Niels Jerne, Georges Köhler, and César Milstein “for theories concerning the specificity in development and control of the immune system and the discovery of the principle for production of monoclonal antibodies.”

Essentially, scientists began demonstrating ways that antibodies against known antigens could be generated in quantity in the laboratory. For example, scientists developed various hybridoma technologies, which involve the fusion of antibody-producing cells—clones of a unique antibody-producing parent cell—with immortalized myeloma tumor cells.

Although hybridoma approaches typically harvest antibody-producing cells from animal models, genetic engineering may be used to produce more human-compatible constructs, namely, chimeric, humanized, or fully humanized monoclonal antibodies. These constructs can avoid the shortcomings of the earliest monoclonal antibodies, which were of murine origin and tended to be immunogenic in humans and to induce disappointingly weak human immune effector responses.

There are other innovative technologies for producing fully human antibodies. In one of these technologies, antibody phage display, human antibody genes are inserted into bacteriophages, which then display the antibodies on their surfaces. After rounds of antigen-guided selection and phage propagation, bacteriophages that express antibodies capable of binding antigen are selected, facilitating the recombinant production of specific monoclonal antibodies.

Making steady progress

In 1986, the first monoclonal antibody was approved for therapeutic use in the United States to prevent the rejection of transplanted organs. The antibody, muromonab-CD3, was of mouse origin, and it was created using hybridoma technology. Because it stimulated the production of anti-mouse antibodies in patients, it caused serious side effects in patients. Eventually, in 2010, it was voluntarily withdrawn from the U.S. market.

As genetic engineering technology improved through the 1990s, opportunities to develop more “human” therapeutic antibodies expanded, and approvals eventually started to tick upward, especially starting in 2014, a year in which the FDA approved seven monoclonal antibodies.

Starting in 2015, developers began focusing more on cancer, driven in part by immune checkpoint modulators. This development does somewhat dilute the novelty of the targeted antigens; that is, many companies are converging on targets such as PD-L1.

Near the end of 2020, when this article was prepared, about 100 monoclonal antibodies for treatment of cancer and noncancer indications had secured FDA approval. “It took a while to gain momentum,” says Janice Reichert, PhD, executive director of the Antibody Society and editor-in-chief of mAbs. “Prior to 2010, a relatively small number of antibodies were in late-stage clinical studies, so there was an insufficient number to drive an increase in approvals.

“Now that more candidates are getting into late-stage clinical studies, a deep pool exists that feeds a steady stream of approvals. I do not anticipate the number of approvals dropping any time soon. Antibodies have a core success rate that seems to be holding over time.”

The global therapeutic monoclonal antibody market was valued at approximately $115 billion in 2018 and is expected to generate revenue of $300 billion by 2025.1 “Interest has been consistent since the inception of monoclonal antibodies,” notes Reichert. “Now that the tools, knowledge, and skill sets are more widely distributed, we are better able to design and implement the therapeutics.

“Monoclonal antibodies, like small molecules, are tools for a specific job. They are an incredibly flexible platform and very good at what they do. It is hard to imagine anything that has the same advantages.”

Unraveling biological complexities

“The greatest challenge now is to understand the biology, not to generate the antibodies,” states Andreas Plückthun, PhD, professor of biochemistry at the University of Zurich. “The whole biology of ligands, receptors, redundancies, antagonists, and so on is much more complicated than it appears.

synthetic designed ankyrin repeat proteins (DARPins)
At the University of Zurich, Andreas Plückthun, PhD, and colleagues use synthetic designed ankyrin repeat proteins c(DARPins) as an alternative to antibodies. In this image, a bispecific DARPin (orange and red) binds to the tumor antigen HER2 (shades of blue), which enters a state in which the generation of growth signals is prevented, bringing the tumor cells to apoptosis. The membrane plane is indicated with a yellow band, and transmembrane helices are shown in yellow.

“In the vast majority of diseases, the body is regulated in a subtle, complicated, and fine-tuned manner. The reason for clinical failures is usually not the technical inadequacy, but rather an inability to adequately understand the biology. Hypotheses might be simplistic or even wrong. One needs to go back to good old mechanistic cell biology and biochemistry and look at the molecular details.”

The idea of delivering a toxic substance to tumor cells or stimulating the immune system around them is not new. But this approach may fail if the antibody does not localize perfectly to the targeted tissue. If the antibody is also active in nontargeted tissues, side reactions narrow the therapeutic window. The localization of these molecules is at best preferential. It is never perfect.

A new research direction of the Plückthun laboratory is to produce the antibody when and where it is needed, essentially inducing the healthy cells, or even the tumor cells, to act as production sites. This approach could lead to platforms that revive previously unworkable ideas and, potentially, induce the production of active effective molecules in relatively small amounts locally.

An antibody library, at its core, is a molecular library where each member binds a target. A particular kind of antibody library has been developed by Plückthun and colleagues, who have noted that protein binders have been generated with various scaffold-based approaches that may serve as alternatives to monoclonal antibody technology.2

The Plückthun laboratory has developed a scaffold-based approach that relies on synthetic designed ankyrin repeat proteins (DARPins). DARPins are very stable and make rapid creation of a biomolecule possible. Therapeutics based on DARPins are in a Phase III trial for macular degeneration; three clinical trials for cancer are at various stages.

In addition, Molecular Partners, a company co-founded by Plückthun, produced an anti-COVID DARPin quickly and inexpensively. The scalable production process could be affordably replicated in other countries, including resource-constrained countries, or the stable molecule could be transported to warm climates. The anti-COVID DARPin has been tested in animal models and found to be highly efficient, and clinical trials commenced in November 2020.

“In academia, we can and should be visionary,” emphasizes Plückthun. “This is our privilege and duty.” Early investigations are underway to rebuild noninfectious viruses by making synthetic molecules inexpensively that could be easily administered by inhalation.

Establishing new approaches

The vast majority of approved antibodies are classical human-type immunoglobulin G1. Although treatment can be very effective, some patients may not respond because they have become refractory or are already resistant to treatment. To improve potency, novel formats such as antibody-drug conjugates, bispecific antibodies (bsAbs), and multispecific antibodies (msAbs) are being used.

Two bsAbs approved for clinical use are blinatumomab, which targets CD19 and CD3 for acute B-cell lymphoblastic leukemia, and emicizumab, which is used to treat clotting disorders (hemophilia A). Over 130 bsAbs are currently in clinical development. To date, more than 20 different commercialized technology platforms are available for bsAb creation and development.3

“The term bsAb is used to describe a large family of molecules designed to recognize two different epitopes or antigens,” says Paul Parren, PhD, executive vice president and head of R&D at Lava Therapeutics. “One of the extremely interesting aspects is that you can design in an obligate activity—a new biology and activity that was not present by just mixing the two antibodies.

bispecific antibodies
Lava Therapeutics develops bispecific antibodies that engage γδ T cells to help them target tumor-expressed antigens such as phosphoantigens. Binding of the γδ T-cell receptor to the butyrophilin-phosphoantigen complex leads to Vγ9Vδ2 T-cell activation and tumor cell killing. The bispecific engagers specifically bind to the conserved T-cell receptor of Vγ9Vδ2 T cells, providing a tumor target–dependent response.

“Approximately 25% of antibodies in development are bsAbs, and about half of those are T-cell engagers targeting cancer. One arm binds the tumor antigen, and the other an antigen on a T cell, bringing the cells together and activating the T cell to kill the tumor cell. This has been a breakthrough.”

But novel technology can lead to novel problems. For a number of targets, this technology creates an exaggerated T-cell activation leading to cytokine release syndrome (CRS) and, therefore, dose-limiting toxicity.

Lava Therapeutics focuses on γδ (gamma delta) T cells, which constitute an effector T-cell population that naturally distinguishes between normal and tumor cells. The company’s γδ T-cell engagers differ in several respects from other T-cell engagers, such as CD3 T-cell engagers. These include increased tumor selectivity and potency (activity against tumor cells) coupled with a lack of a risk for CRS. Preclinical data are very promising in nonhuman primates and patient tumor materials. Lava’s lead program is expected to enter Phase I/IIa trials in the first quarter of 2021, and the company expects to have an additional program in the clinic by the third quarter of the year.

“We have a whole proprietary toolbox available now to modify antibodies to tailor them for the functionality we want for innovative therapies,” Parren adds. “Bispecifics and multispecifics have a big future, and not just at Lava.” For example, Sanofi recently moved the first trispecific antibodies to trials.

Enabling research

For decades, scientists produced monoclonal antibodies by fusing the B cells that produce antibodies of interest with immortalized cells to create hybridomas, perpetual antibody factories. A landmark step in the late 1990s was the development of display technologies in which organisms such as yeast or bacteria are engineered to express libraries of antibodies or antibody fragments. A more recent development, single-cell sequencing, allows identification of a gene that encodes a specific antibody.

“These different methods are not mutually exclusive and have transformed our ability to make, understand, and profile monoclonal antibodies,” says Alex Burgin, PhD, executive director at the Institute for Protein Innovation (IPI), a nonprofit research organization.

“The future,” he insists, “is fully realizing what antibodies can do beyond binding and blocking, bringing things together, for example, or locking a protein or target in a particular combination, or being able to discriminate different cellular states.”

The ability to make antibodies that bind specific protein conformations requires expressing and purifying certain protein targets, such as complex cell surface receptors. Burgin notes, “You need scientists dedicated to understanding protein function and structure if you are to make progress.

“Getting antibodies to recognize different conformations is very challenging when using traditional immunization methods. It is very difficult to control what the immune system sees as it tries to recognize a foreign protein. In display approaches, we can control the protein and lock it into the desired targeted conformation.”

Another important advantage is that display technologies can be used to discover antibodies that target highly conserved proteins between species. These cannot be made through traditional immunization because the immune system is programmed to prevent self-recognition.

The IPI has developed a highly automated, high-throughput pipeline and has capabilities to make hundreds of antigens simultaneously. It has also built a robust yeast antibody display platform. The IPI’s libraries, automation, and ability to parallel process many antigens are unique. “Our goal,” Burgin explains, “is to make antibodies to target every protein in the extracellular and secreted proteome— thousands of proteins, many of which are highly conserved and difficult to make.

“We are also very enthusiastic about generating antibodies that will recognize different glycosylation, phosphorylation, or conformational states. As a nonprofit, our mission is to enable others. We can operate at a scale that is often unattainable by a traditional academic laboratory. Our focus is to provide research reagents to accelerate science, and if that results in new therapeutics, that is fantastic.”

Ensuring quality

Monoclonal antibodies are used extensively in diagnostic tests and basic research in addition to the therapeutics that have dramatically changed the landscape of medicine and human health. Their tremendous effects are due to their specificity and sensitivity to detect proteins or other targets of interest. A vast number of detection technologies use antibodies by adding a variety of moieties.

“We were the first company to develop and commercialize antibodies to detect phosphorylation modifications, and that opened the door to study cell signaling,” asserts Roberto Polakiewicz, PhD, CSO, Cell Signaling Technology. “Now, we focus on cancer, neuroscience, developmental biology, metabolism, and other areas of biology that overlap with disease areas.”

No single assay can validate an antibody, as these materials from Cell Signaling Technology demonstrate. Two antibody clones directed against target 53BP1 perform as expected on control and knockout cell lysates, but not on tissue sections. Left section: The clone stains the nucleus (as expected). Right section: The clone stains the nucleus and some cells at the membrane (not as expected). This result could not have been predicted from the western blot data, highlighting the need for application-specific antibody validation.

The company’s scientists have expertise in several fields. They perform their own research and also collaborate with key opinion leaders and an expansive scientific network. Using in-house bioinformatics capabilities, they can parse through information and inform decision making for new products.

Polakiewicz notes that the main challenge in the research market is quality—specificity, sensitivity, robustness, and reproducibility. “We are at the forefront and the top recognized company for quality,” he maintains. “Both our monoclonal and also our polyclonal antibodies are extremely well validated. Unfortunately, many other available tools are not properly validated, and that results in repeating experiments or even, unfortunately, inaccurate results.”

Validation requires a full range of experiments to show the antibody’s specificity (in detecting the target and nothing else under the conditions of the experiment) and sensitivity (in detecting targets that are present at low levels).

“In the future, monoclonal antibodies will have expanded use over polyclonal versions,” Polakiewicz continues. “There will be more recombinant antibodies, more extensive engineering of function for different applications, and more hybrid molecules with antibodies and DNA attached to them to enable different types of assays.

“I see a very rapid expansion of this field simply because antibodies are fantastic and the best tool for detection and building assays. Mother Nature engineered these beautiful things for hundreds of thousands of years. We are pretty smart, but we are not smarter than nature yet.”

References
1. Lu RM, Hwang UC, Liu IJ, et al. Development of therapeutic antibodies for the treatment of diseases. J. Biomed. Sci. 2020; 27: 1. DOI: 10.1186/s12929-019-0592-z.
2. Wu Y, Batyuk A, Honegger A, et al. Rigidly connected multispecific artificial binders with adjustable geometries. Sci. Rep. 2017; 7(1): 11217. DOI: 10.1038/s41598-017-11472-x.
3. Labrijn AF, Janmaat ML, Reichert JM, Parren PWHI. Bispecific antibodies: a mechanistic review of the pipeline. Nat. Rev. Drug Discov. 2019; 18(8): 585–608. DOI: 10.1038/s41573-019-0028-1.