Building an All-Human Antibody Against Cancer Through an In Vitro Germinal Center
Scientific fashion changes. Currently "out" is basic science research into arcane mechanisms, and "in" is translational research that promises relatively quick clinical reward. The push for more translational research is a very laudable one—it is meant to speed treatments to patients who desperately need them. And yet, as shown by the example of Natalie Sutkowski, PhD, Assistant Professor in MUSC’s Department of Microbiology & Immunology, we must be sure that this drive for clinical application does not squelch scientific curiosity and serendipity and realize that sometimes the most fertile ground for therapeutic innovation lies where we might least expect it.
What, for instance, could we learn from a now highly defective retrovirus inserted randomly into the human genome many millennia ago? What possible clinical relevance could it have?
And yet it piqued the curiosity of Dr. Sutkowski. Why would this random insertion have been retained? What purpose could it be serving? The answer to those questions led to a critical methodological breakthrough that will speed the development of human monoclonal antibody therapy for cancer patients. The seemingly most arcane of interests led to a new method of producing greater quantities of better-quality and more clinically effective antibodies (longer half-life, tighter binding capacity)—a technology so promising that in 2010 MUSC licensed the rights to the technology to the Charleston-based biotechnology startup, Immunologix, Inc, which was acquired in 2011 by Intrexon Corporation, a Virginia-based synthetic biology company. All of this because superantigens, and a vestigial retroviral superantigen encoded in the human genome in particular, aroused Dr. Sutkowski’s curiosity.
What Is a Superantigen?
Unlike a typical antigen, which binds on average to only about 1 in every 10,000 naïve T cells, superantigens can bind to at least 1 in 50 (even more if redundancy is factored in), meaning that the T cell response elicited by these pathogen-derived proteins is far stronger.
Each T cell has a unique T cell receptor. A close match is needed by a typical T cell receptor for antigen binding to occur. The large number of T cell receptor genes in our genome helps ensure a diverse immune system that can mount a defense against almost any antigen. The immune system must come up with just the right combination in order to bind to a pathogen and counter its attack. Conventional antigens bind to a unique combination of five T cell receptor variable gene products. In contrast, a superantigen need bind only to the V-β gene product, thus greatly increasing the chances of a good fit.
Why then is a superantigen, the retrovirus HERV-K18 envelope protein, encoded in the human genome? It was introduced into the genome relatively late in evolutionary terms, after the continental split but before the advent of humans, as evidenced by its being found in the genetic code of Old World but not New World monkeys. Although now very defective, as would be predicted for genetic code introduced millennia ago, its envelope protein is still functional and capable of eliciting a strong T cell response. Its insertion was probably random, but why has it been retained?
Dr. Sutkowski surmised that, for the superantigen to have been retained so long in the genetic code, it had to be associated with some sort of survival benefit. But how could the strong T cell response elicited by the superantigen ensure its survival?
Enter Epstein-Barr Virus
Dr. Sutkowski thought she had an answer: Epstein-Barr Virus (EBV), the virus that infects tonsil B cells and causes mononucleosis, turns on the superantigen. Almost all adult humans (>90% worldwide) carry the virus. In the early stage of infection, EBV has 8 oncogenes that cause these B cells to divide endlessly and are associated with B cell cancers (ie, AIDS-related and post-transplant lymphomas). They also stimulate the superantigen and consequently T cells. Tonsil T cells attack EBV-infected B cells expressing the oncogenes. However, EBV escapes the T cells by turning off its oncogenes as the B cells differentiate into memory cells when the EBV-infected B cells reach the bloodstream. EBV becomes dormant and goes into hiding, content to live out its days quietly in the host in memory B cells.
The EBV infection process mimics the body’s natural ability to generate antibody immune responses. Naïve B cells respond to antigen early during an infection and produce IgM during the initial response. As those B cells then become experienced, they are able to recognize the target more easily. This transformation, along with the immunoglobulin class switch from IgM to IgG production, typically occurs in the germinal center of a lymph node. A T cell response is required for the naïve B cells to differentiate into experienced (memory) B cells in the germinal center.
Dr. Sutkowski’s thought was that EBV could be using the strong T cell response elicited by the superantigen to encourage the differentiation of EBV-infected naïve B cells into memory cells, where EBV would persist for the lifetime of the host, thereby promoting its long-term survival. Moreover, B cell antibody responses might be generated in conjunction with superantigen-mediated T cell activation.
A Light Bulb Moment
It’s then that the light bulb switched on.
Could EBV’s cooptation of the strong T cell response elicited by the superantigen provide a model for building an antibody response ex vivo? Could one devise an in vitro germinal center (Figure 1) promoting the differentiation of naïve B cells into experienced ones, with class switching from IgM to IgG antibodies?
FIGURE 1. An in vitro germinal center. Ag = antigen; EBV = Epstein-Barr virus; Ig = immunoglobulin. Medical Illustration by Dr. Natalie Sutkowski and Sargent Illustration and Design, LLC.
Dr. Sutkowski thought so. She took tonsil B cells and transformed them overnight with EBV, essentially intending to make them into immortalized cancer B cells that would continuously produce antibodies. The extent of transformation could be gauged with green-fluorescent protein–labeled virus that would cause infected B cells to glow green. By concentrating the virus and using a spinoculation technique, almost 100% of the cells glowed green overnight, demonstrating exceptional infection. Dr. Sutkowski and her team, including Semyon Rubinchik, PhD, were then ready to look at B cell differentiation of the EBV-transformed cells.
With EBV transformation alone, most of the B cells produced IgM antibodies into their cell supernatant. However, when EBV transformation was accompanied by the two main differentiation signals found in a germinal center—antigen binding and T cell help—class switching to IgG ensued, and IgG antibodies were then secreted into the cell supernatant. The germinal center signals mimicked antigen binding to the B cell in the presence of a “T cell bath,” which represents the soluble factors produced by antigen-activated T cells in the germinal center (or alternatively by superantigen-activated T cells). Although the mechanisms are still under investigation, it appears that, after the “T cell bath,” class switching occurs, and the IgG genes continue to be refined, which could lead to increased antibody affinity (a process known as somatic hypermutation).
In short, fully human IgG antibody of high affinity was produced in vitro.
Dr. Sutkowski had in effect created an in vitro germinal center that could be used to generate immortalized B cells producing large quantities of antibody to many antigens. The antibodies are largely of the IgG class, which can be used in the clinic because they have a much longer half-life in the body compared with IgM and generally a much higher affinity for antigen binding. From a technological standpoint, sandwich ELISA, Western blotting, or other means of high-throughput screening could then be used to isolate B cells producing antibodies of interest for clonal expansion (Figure 2), and monoclonal antibodies would be isolated from the cell supernatant.\
Curiosity had led to a breakthrough in methodology that removed many of the barriers to clinical application of monoclonal antibody therapy.
FIGURE 2. A novel method for producing all-human monoclonal antibodies. EBV = Epstein-Barr virus; HuMAbs = human monoclonal antibodies.; IgG = immunoglobulin G. Medical Illustration by Dr. Natalie Sutkowski and Sargent Illustration and Design, LLC.
Why Is This Technology Better?
The antibody technology developed by Dr. Sutkowski allows for the in vitro production of fully human antibodies and does not involve animals. This cuts down on time and expense, and most importantly, obviates ethical issues. Traditionally, monoclonal antibodies have been produced by injecting antigens into mice, but the resulting mouse antibodies could not be injected into humans because they set off an immune reaction known as serum sickness. Furthermore, mouse-derived monoclonal antibodies had a much shorter half-life in the body, usually being cleared within a day or two, limiting their efficacy.
In an effort to overcome these barriers, scientists engineered chimeric antibodies—part mouse and part human. Mouse antibodies were harvested, and then a part of that antibody (the Fc region associated with half-life) was removed and replaced by its human counterpart. The chimeric antibody could then use the FAb section (of mouse origin) to target antigens but would last longer in the body due to the human Fc component. These chimeric antibodies were clearly more attractive for therapeutic applications than purely mouse ones because they remained in the system long enough to have some of the intended effect against the target. And yet considerable time and expense were required for their manufacture and the possibility of an immune reaction to the mouse components of the chimera remained.
Another approach was to develop a xenomouse, whose own immune system had been completely replaced with a human one. Scientists lined up to inject their antigen into this patented mouse model to generate human antibodies, but the line grew so long that much research into antibody therapy stalled as a result.
With Dr. Sutkowski’s technology, no mouse or any other animal is required. Fully human antibodies can be produced in the laboratory, and the antibodies are of the IgG type and so are better able to target antigens and to last longer in the body.
This technology can be used to generate antibodies not only against any cancer but also against infectious diseases or toxins. Because it does not depend on a laboratory animal for production, antibodies against the most lethal of infections or toxins can be tested (ie, there is no danger of the animal dying from the infection or toxin before the antibody is produced).
Most importantly, the technology can be used to create therapeutic antibodies against self-antigens. Most of the therapeutic antibodies in the clinic today target growth factors and their receptors, as well as tumor antigens, which are overexpressed in tumor cells compared with normal cells. Since these antigens are part of our body, the therapeutic antibodies target ‘self.’ While low-affinity, IgM-producing self-reactive B cells are often found in the body, in general B cells producing high-affinity IgG antibodies are not made so that we do not develop autoimmunity. However, such B cells might protect against cancer and other diseases by making protective antibodies. With the use of Dr. Sutkowski’s approach, the low-affinity IgM producers could be immortalized and converted in vitro to IgG producers, allowing for the production of self-reactive antibodies.
This was quite simply a giant step forward in the development of monoclonal antibodies for clinical application, and it resulted from the curiosity of a single investigator about a mechanism that many would have passed over because it seemed to have little clinical usefulness or to be of purely “academic” relevance.
Developing a Better Antibody Against Breast and Other Cancers
As proof of concept that the technology could generate therapeutic antibodies, Dr. Sutkowski teamed up with Daniel Fernandes, PhD, DSc, Professor in MUSC’s Department of Biochemistry and Molecular Biology, and Associate Co-Director of Translational Research at the Hollings Cancer Center, to develop antibodies against a tumor antigen called nucleolin. Long before Dr. Sutkowski came to MUSC, Dr. Fernandes had identified nucleolin as a good target for anti-cancer treatments. Nucleolin is an important protein found in the nucleus of all normal cells. But by comparing cancerous and noncancerous cells, Dr. Fernandes discovered that in the former nucleolin was not restricted to the nucleus but occurred throughout the cytoplasm and on the cell surface, where it could be targeted with an antibody.
After hearing about her antibody technology, Dr. Fernandes sought out Dr. Sutkowski to work jointly to develop anti-cancer antibodies targeting nucleolin.
Together they have received a Department of Defense grant to develop targeted immunotherapy against breast cancer using a human antibody against nucleolin (Figure 3) and to try to bring it to clinical trials within two years. Herceptin is already available to target breast cancer, but it can only target certain types. Because nucleolin is expressed on the surface of almost all breast cancer cells, it could show efficacy against a much broader range of breast cancer types. Dr. Fernandes has also studied the potential application of an antibody against nucleolin to the treatment of chronic lymphocytic leukemia, as well as other cancers.
FIGURE 3. Selective antibody targeting of breast tumor cells. Medical Illustration by Dr. Natalie Sutkowski and Sargent Illustration and Design, LLC.
So far, targeted therapy using an antibody against nucleolin is showing impressive results. In studies performed by Wei Sun, PhD, in Dr. Sutkowski’s laboratory, virtually all breast cancer cells are dead after 96 hours of treatment in vitro. Currently, Drs. Sutkowski and Fernandes are testing this therapy in murine models to see if the results can be replicated there and whether there is any rebound of cancer cells as time passes after initial treatment. If results are as promising there, the next step would be an investigational new drug application to the US Food and Drug Administration and then a phase 1 clinical trial carried out at MUSC’s Hollings Cancer Center. Hollings is one of only 67 National Cancer Institute–designated cancer centers in the nation.
Nucleolin may well prove a very effective target for anti-cancer immunotherapeutic approaches. If so, a fully human nucleolin-targeted antibody will have greater clinical promise than a mouse-derived or chimeric one because it is far less likely to elicit a problematic immune response.
The ability to produce a fully human antibody has paved the way for monoclonal antibodies to come into their own as important new additions to the cancer clinician’s armamentarium.