Theo Rispens is group leader at the department of Immunopathology Research. His research topics cover all aspects of antibodies, from structure-function analysis of IVIg to regulation of antibody production. This includes immunogenicity of biologicals.
You have studied ‘hard-core’ chemistry, what brought you here at Sanquin?
“My PHD project involved research in the field of physical chemistry. Although the science was interesting, I missed a connection with society. At Sanquin I got the opportunity to combine basic chemistry with translational, patient-directed research. I started with a project that was fairly physical-chemical in nature, studying the stability of therapeutic immunoglobulin preparations and developing methods to do so. Within Sanquins bio-medical work field, my background in chemistry helps me to perceive biological processes from a molecular point of view. I try to imagine how fast individual molecules move about and how they interact with each other. An immune response is all about recognition of dangerous pathogens, which ultimately involves interaction and binding of molecules, like antibodies to antigens.
What is so fascinating about antibodies?
“Antibodies are molecules that show endless structural variations. But what they all have in common is their bridging capacity. Antibodies span bridges between pathogens and a variety of immune cells and thereby one way or another activate our immune system. What intrigues me most is how B-cells randomly combine a set of gene fragments to produce this enormous diversity of antibodies that can fulfill a variety of functions. The immune system is able to generate antibodies against virtually any thinkable antigen. Depending on their classes or subclasses or glycosylation, antibodies are even capable to dampen the immune reaction, like for instance IgG4 does. Sometimes we produce harmful antibodies, as we see in autoimmune diseases. In the clinic antibody-based therapy has become established these days. In various medical fields a wide variety of therapeutic antibodies is being used, all with their own unique features.
What makes IgG4 so special?
Antibodies have two identical Fab arms that bind antigen. However, “IgG4 exchanges half-molecules in the body, and that way obtains two different Fab arms. Therefore, unlike IgG1, IgG4 binds to its target with only one Fab-arm, cannot crosslink and therefore cannot form large immune complexes. That’s one reason why IgG4 is a poor trigger of any effector function. It is less effective in activating the rest of the immune system. You could see IgG4 as a mild antibody. IgG4 can block the antigen, but not trigger further action. When we are exposed to an antigen for a long time, for instance to a food antigen, the immune system often switches to IgG4 production, which may dampen the immune response and prevent overacting by the immune system. This is another form of regulation of the immune response.
However, sometimes IgG4 is associated with disease, like in IgG4 related disease. This disease has only been known for 10-15 years. Many organs can be affected, including the pancreas; this type of pancreatitis is often mistaken for a tumor. The pancreas is full of IgG4-producing plasma cells, and serum IgG4 titers are often extremely high. Treatment usually consists of corticosteroids. I’d like to know whether IgG4 is the cause or the result of this disease. What is the mechanism behind this disease? Why is the body making so much IgG4, is there a particular antigen?”
How does glycosylation affect the functioning of antibodies?
“All antibodies have a glycosylation site in the Fc-part of the molecule, the tail that is involved in effector functions such as complement activation and binding to Fc-receptors on immune cells. Together with other departments of Sanquin we try to find out how various glycans affect effector functions. How and when does glycosylation of antibodies take place in our bodies? Can we design antibodies with more potent effector functions? This could be of interest for development of therapeutic antibodies. Besides Fc-glycosylation, some antibodies have glycosylation sites on the Fab–arm, depending on the amino acid composition. We know very little about this Fab glycosylation. It is feasible that Fab glycosylation could affect the binding to the antigen. We’d like to find out whether this glycosylation varies in the course of the immune response and whether it contributes to regulating this response, or may contribute to affinity maturation.”
Why do some patients make antibodies to biologicals and others don’t?
“I would like to turn the question around. Why do not all patients make antibodies to their therapeutics? A growing number of patients is treated with therapeutic antibodies and derivative molecules, for a variety of diseases. Depending on the test we use we are able to detect antibodies to some of these drugs in a majority of patients. Regardless how “human” a therapeutic antibody may be, it will always include a foreign epitope, the antigen-binding site, the idiotype, which could elicit an immune response. We think genetic components may determine whether patients produce only sub-clinical amounts of antibody, or develop a full-blown immune response, completely inhibiting their biological. It is impossible to predict immunogenicity at this moment. What we can do is optimize dosage and dosing schedule to overcome immunogenicity. Co-medication of immune suppressiva is also important in reducing the immune response.
An anti-idiotype response itself is not unheard of. All the time we generate new antibodies that our immune system hasn’t seen before. Niels Jerne won the Nobel prize in part for his antibody-network theory. He claimed that our body makes antibodies to our own antibodies. And subsequently antibodies to those antibodies. And so on. With as final result antibody responses that are regulated by (auto-)antibody responses. Because therapeutic antibodies are directed to a human target protein, the antibody response to a biological can also be interpreted as an attempt of the immune system to neutralize these autoantibodies.
How would you develop a biological that is not immunogenic?
“It is not necessary to completely eliminate immunogenicity, as long as it remains low and doesn’t affect therapy. I think a good approach would be to remove dominant T-cell epitopes from therapeutic antibodies. Though prediction of those epitopes is not flawless, we would already make a lot of progress if the worst cases would be eliminated. T-cell reactivity is something that you can test in a lab, by exposing peptides to panels of cells of various people. Furthermore, choosing antibody clones with high stability will minimize aggregation of the biological, which is a known cause of immunogenicity. If the function of a therapeutic antibody is to block a target, IgG4 would be a good choice. In theory, such an antibody might be less immunogenic than its IgG1 counterpart, because it has less options to interact with the immune system. However, this has not yet been proven.