Engineering Better Agonistic Antibodies for Cancer Immunotherapy
Agonistic antibodies have shown potential in preclinical studies, but translation depends on optimized engineering.
Antibodies have been applied therapeutically since the 19th century, predating the widespread scientific use of the term “antibody.”
At the time, scientists were seeking a treatment for diphtheria, which had caused several major epidemics. They discovered that the sera of surviving animals contained neutralizing agents that cured the disease when transferred to infected animals. This led to the first therapeutic serum against diphtheria.
A few years later, Dr. Paul Ehrlich worked with that same serum. His work provided the foundation for several concepts that are fundamental to modern immunology. Ehrlich described the neutralizing agents as “antibodies” and proposed that cells carry structures that, when bound by such molecules, induce specific cellular effects.
Figure 1: A brief history of the antibody. Adapted from “Introducing the Antibody.” Credit: Technology Networks.
These ideas have been refined over the years, but the basic principles were correct. Eventually, hybridoma technology facilitated the production of consistent therapeutic antibodies, and today, antibodies have been successfully applied across various disease indications. Much of this success can be attributed to antagonistic antibodies, which bind a specific target to inhibit its biological activity. Meanwhile, the development and translation of agonistic antibodies, designed to stimulate cellular activity, has lagged.
Technology Networks spoke with Dr. Mark Cragg, professor of experimental cancer biology at the University of Southampton, to learn more about the challenges and opportunities surrounding agonistic antibodies. Cragg has spent over 30 years studying antibody biology, and his pursuit of better therapeutics is driven by personal experience.
“My interest began with a curiosity about biology and how living systems work,” said Cragg. “Cancer then affected my family, turning that curiosity into more personal questions: what is cancer, where does it come from, and why does it happen?”
These questions motivated Cragg to seek answers, with a focus on cancer research.
What makes some antibodies more effective than others in driving tumor regression?
Mark Cragg, PhD (MC):
The most effective antibodies are those whose properties are matched to the biological mechanism needed to eliminate the tumor.
Anti-CD20 antibodies illustrate this well: although they can trigger several fragment crystallizable (Fc)-dependent and Fc-independent effects, macrophage-mediated phagocytosis is especially important in relevant mouse models. Type 2 [anti-CD20] antibodies, such as obinutuzumab, can perform better because they remain on the cell surface for longer and sustain phagocytosis over time.
Obinutuzumab
Obinutuzumab is licensed for the treatment of chronic lymphocytic leukemia and follicular lymphoma, as well as kidney and autoimmune conditions. It is a type 2 antibody and has been shown to undergo less internalization upon binding CD20 than type 1 antibodies, which may explain its higher potency. Type 1 and 2 anti-CD20 antibodies differ in how they bind the target receptor.
Efficacy also depends on context. Tumors contain different immune-cell populations and activation states, and repolarizing a suppressive tumor microenvironment can turn a resistant model into a responsive one.
Restimulating tumors with agonistic antibodies
Immunostimulatory agonistic antibodies target receptors on immune cells or antigen-presenting cells. They enhance the anti-tumor immune response either by activating the immune system itself or by increasing the tumor's susceptibility.
Finally, antibody geometry matters. Epitope location and orientation can strongly influence antibody function. For some targets, membrane-proximal binding favors Fc-dependent cellular effector mechanisms, whereas particular membrane-distal epitopes can support the receptor geometry required for agonistic signaling.
“The optimal position is target- and mechanism-dependent.” — Dr. Mark Cragg.
IH:
What makes agonistic antibodies interesting, and how do factors such as isotype, epitope, hinge flexibility, and affinity influence how effectively they activate a receptor?
MC:
Agonistic antibodies are compelling because, in preclinical models, stimulating an immune receptor can match—and in some settings exceed—the activity of checkpoint blockade.
The clinical challenge is achieving the right degree of receptor clustering in the right place: insufficient clustering produces weak activation, whereas excessive or systemic activation can cause on-target, off-tumor effects and cytokine-mediated toxicity.
Receptor cross-linking and clustering
Some therapeutic antibodies rely on cross-linking or clustering of receptors to trigger a response. In cross-linking, antibodies bring together two or more receptor molecules into proximity. In receptor clustering, many receptors are brought together as a result of cross-linking. This enables the receptors to interact and initiate signaling. If interactions are insufficient, the desired therapeutic signaling may be inadequate; if interactions are excessive, toxicity may occur.
Several antibody properties govern this balance:
- Epitope location influences receptor geometry; membrane-distal binding can favor clustering.
- Isotype and Fc structure—including glycosylation and engineered Fc variants—influence the pattern and strength of Fc gamma receptor engagement, while differences between mouse and human Fc receptor biology can limit translation.
- Hinge flexibility also matters. Increasing rigidity can hold receptors in a configuration that supports stronger signaling.
- Affinity contributes, but binding strength alone is not enough; effective agonism depends on how the antibody organizes receptors at the cell surface.
IH:
Agonistic antibodies haven't yet had the same clinical success as checkpoint blockers. What do you think has held them back, and what could help change that?
MC:
Checkpoint blockers are comparatively straightforward to design: they just need to bind the target with sufficient affinity to prevent an inhibitory interaction. Agonistic antibodies face a much more complex task. They must bind in a geometry that brings receptors together in the correct configuration to generate the required downstream signal, while avoiding activation in healthy tissues.
Affinity, epitope, isotype, hinge properties, and engagement of other receptors can all influence that outcome. Translation is further hindered by incomplete knowledge of how receptors are organized and move within the membrane. That is, whether they exist as monomers or preformed assemblies, and how that organization changes on different cell types and after binding.
AI is becoming increasingly effective at proposing antibody sequences, structures, and molecular interactions for experimental testing, but it cannot yet replace experiments that test the relevant biology. Progress will therefore depend on combining rational antibody engineering with human-relevant systems that measure receptor organization, signaling, efficacy, and toxicity.
IH:
What’s one emerging trend or technology in antibody discovery that you think could have a major impact in the years ahead?
MC:
I feel the major advance will be better human-relevant functional assays.
“The bottleneck is no longer simply making antibodies; it is deciding which antibodies to make, what biological effect they should produce, and which candidates are most likely to work in patients.” — Dr. Mark Cragg.
For immunostimulatory antibodies, for example, multivalent formats may improve receptor clustering, but we still need to establish whether they reproduce a productive signaling configuration and retain activity in the presence of clinically relevant barriers, including an immunosuppressive tumor microenvironment and dysfunctional or exhausted T cells.
AI should accelerate antibody design and help propose antibody structures and molecular interactions for experimental testing. Its clinical value, however, will depend on high-quality experimental systems that capture human biology. Rather than screening ever-larger numbers of molecules using limited readouts, the field needs assays that reveal which candidates deliver the intended function in a clinically relevant context.
Add Technology Networks as a preferred Google source to see more of our trusted coverage.
