April 2025: Affinity Maturation

Affinity maturation is a fundamental process in the adaptive immune response that results in the selection of B cells producing antibodies with increased binding affinity for their target antigen. This process occurs in germinal centers and is driven by somatic hypermutation and selection pressures. Understanding the mechanisms that regulate affinity maturation has significant implications for therapeutic antibody development and vaccine design. This past month, several recent studies have explored both the biological underpinnings of affinity maturation and computational methods to engineer antibodies with improved binding properties.

Biological Regulation of Affinity Maturation

A recent study investigated how monoclonal antibody (mAb) infusion during immunization influences germinal center dynamics and B cell selection. Mice were immunized with HIV gp120 and infused with CD4 binding-site-specific mAbs of varying affinities. The results showed that while mAb infusion reduced SHM and affinity in most competing B cells, a subset exhibited increased SHM and affinity. High-throughput sequencing of plasma cells revealed a shift in phylogenetic tree topology, suggesting more rapid differentiation. This study highlights that high-affinity mAbs can suppress the recruitment of low-affinity clones while enhancing selection pressure, accelerating affinity maturation in specific epitope-targeted B cells.    

Another study examined how antibody affinity influences the transition of B cells into the plasma cell compartment. Using mouse models, it was found that plasma cell precursors expressing high-affinity antibodies received more T follicular helper cell (TFH) support and underwent more rapid division than their lower-affinity counterparts. These findings provide insight into how serological affinity maturation occurs post-germinal center, emphasizing that plasma cell output is not solely based on germinal center affinity but also on downstream affinity-sensitive mechanisms.

Meanwhile, Merkenschlager et al.’s work tested a theoretical model predicting that high-affinity B cells might reduce their mutation rate per division to limit deleterious mutations. In mice immunized with SARS-CoV-2 antigens or model antigens, high-affinity B cell clones shortened their G0/G1 phase and exhibited reduced SHM per division. This adaptive mechanism preserves beneficial mutations and minimizes the risk of affinity loss in dominant clones. The study provides a refined view of how mutation rates are modulated to optimize affinity maturation outcomes.    

Structural and Mutational Basis of Affinity Enhancement

One aspect of affinity maturation studied by researchers at the Icahn School of Medicine at Mount Sinai, is the enrichment of clonotype-specific mutations in public antibodies. In an analysis of SHM patterns, researchers identified mutations in a model IGHV4-59/IGKV3-20 antibody that were highly enriched in a public clonotype targeting a specific viral epitope. These mutations, which appeared more frequently than in other antibodies using the same V genes but recognizing different epitopes, were shown to significantly enhance binding. This suggests a degree of convergence in affinity-enhancing mutations across individuals, reinforcing the potential for vaccine strategies that stimulate the generation of such public antibody responses.    

Beyond mutation enrichment, structural adaptations also play a vital role in optimizing antigen binding. Comparative structural analysis of two affinity-matured scFv antibodies (C6 and E11) revealed key mutations (such as Lys58 to Arg58 in CDR2 and the introduction of a disulfide bond in CDR3) that significantly enhanced binding affinity. Interestingly, thermodynamic analysis demonstrated that while affinity maturation increased structural flexibility and reduced stability, these changes ultimately improved antigen recognition. This trade-off between flexibility and binding strength highlights an important principle in antibody evolution: structural dynamics can enhance function, even at the cost of reduced intrinsic stability.

Building on these principles, antibody engineering efforts have successfully applied affinity maturation strategies to develop high-affinity therapeutics. Targeting a cryptic class 6 epitope on the SARS-CoV-2 receptor-binding domain (RBD), researchers in Australia used in vitro display technology to drive affinity maturation, yielding antibodies with low picomolar affinity and potent neutralization against variants of concern. Structural studies identified binding sites that effectively avoided mutational hotspots, further enhancing therapeutic potential. Notably, class 6-targeting antibodies were found in human memory B cells and could be induced in transgenic mice, emphasizing their relevance for long-term immunity and pandemic preparedness.    

Computational and Experimental Engineering of Antibody Affinity

Recent work has highlighted multiple strategies for enhancing antibody affinity through computational modeling and experimental screening. The following studies exemplify complementary approaches—deep learning and sequence-based generative modeling.

Two models, GearBind and AffinityFlow, demonstrate how machine learning can accelerate affinity maturation without the need for iterative wet lab cycles.

• GearBind uses a geometric graph neural network to model antibody-antigen interactions.

• Employs contrastive pretraining on large structural datasets

• Improved binding observed in two distinct antibody formats

• Reduced EC50 and KD values validated the enhanced affinity

• AffinityFlow takes a structure-free approach by integrating:

• AlphaFlow-based structure generation

• Alternating optimization between sequence and structure predictors

• A co-teaching module to manage noisy training signals

• Outperformed baseline models in generating high-affinity variants using only sequence data

Together, these tools reflect a shift toward efficient, model-driven antibody design with potential for broad therapeutic use.