frontier-banner
前沿速递
首页>前沿速递>

Nature Communications | High-Throughput Engineering of Bispecific Antibodies Enhances Macrophage-Mediated Cytotoxicity Against B-Cell Lymphoma

Nature Communications | High-Throughput Engineering of Bispecific Antibodies Enhances Macrophage-Mediated Cytotoxicity Against B-Cell Lymphoma
--

This study systematically identified multiple targetable B-cell lymphoma surface antigens using a 'tactical surfaceome profiling' strategy, providing functionally validated candidate target combinations for developing novel macrophage-directed immunotherapies, particularly offering new insights into overcoming resistance to conventional CD20 monoclonal antibodies.

 

Literature Overview

The article titled 'High-throughput engineering of bispecific antibodies to enhance macrophage-mediated cytotoxicity of B-cell lymphoma,' published in Nature Communications, systematically explores how high-throughput screening and engineered bispecific antibodies can activate macrophages to enhance killing of B-cell lymphoma. The research team developed a strategy called 'tactical surfaceome profiling,' combining functional screening with antibody engineering to identify several new targets that effectively mediate antibody-dependent cellular phagocytosis (ADCP). The authors further constructed 156 bispecific antibodies and validated their anti-tumor activity in both in vitro and in vivo models. Ultimately, a bispecific antibody incorporating a low-affinity SIRPα decoy domain and an anti-CD38 arm (WTa2d1xCD38) demonstrated superior efficacy and safety. This study provides a new pathway for immunotherapy of B-cell non-Hodgkin lymphoma (B-NHL), particularly in contexts where T-cell function is impaired or MHC-I is absent, making macrophages key effector cells.

Background Knowledge

1. The challenges in B-cell lymphoma addressed by this study: Although rituximab (anti-CD20) combined with chemotherapy (e.g., R-CHOP) has significantly improved outcomes for patients with diffuse large B-cell lymphoma (DLBCL), 30–40% of patients still experience primary or secondary resistance. Resistance mechanisms include loss of MHC-I, insufficient T-cell infiltration, and activation of immune checkpoints within the tumor microenvironment. Moreover, over 50% of B-cell lymphomas exhibit immune escape mechanisms from T-cell recognition, limiting the efficacy of CAR-T or T-cell engagers. Therefore, targeting the innate immune system, particularly macrophages, has become an important alternative strategy.
2. Current research bottlenecks with CD47: The CD47/SIRPα pathway delivers a 'don't eat me' signal to macrophages, and blocking this pathway enhances ADCP. However, anti-CD47 monoclonal antibodies or SIRPα fusion proteins often cause severe anemia and thrombocytopenia in clinical settings due to widespread expression on red blood cells and platelets. High-affinity anti-CD47 antibodies result in significant on-target toxicity, limiting their therapeutic window.
3. Research rationale: The authors propose a 'function-first' screening strategy, avoiding preconceived target selection, instead directly evaluating each antibody's ability to activate macrophage phagocytosis of lymphoma cells using a comprehensive surfaceome antibody library. By combining cross-species validation (mouse and human systems), they identified conserved and functionally active targets. Furthermore, through bispecific antibody engineering, they combined 'eat me' signals (e.g., anti-CD38) with 'block don't eat me' signals (e.g., SIRPα decoy), achieving synergistic macrophage activation while reducing toxicity risks. This approach precisely circumvents the high toxicity associated with CD47 targeting by replacing traditional anti-CD47 antibodies with a SIRPα decoy and exploring the use of low-affinity variants (WTa2d1).

 

 

Research Methods and Experiments

The authors first established a high-throughput co-culture system using M-CSF-differentiated primary mouse and human macrophages co-cultured with fluorescently labeled B-cell lymphoma cells (e.g., A20, Raji, Toledo). They added arrayed monoclonal antibody libraries (173 mouse antibodies, 241 human antibodies) and evaluated tumor cell clearance under monotherapy, combination with anti-CD47, or anti-CD20 conditions. Time-lapse microscopy was used to quantify changes in fluorescence area to identify antibodies that significantly promoted phagocytosis. Flow cytometry was also employed to assess antibody binding strength, revealing only a weak correlation between binding levels and functional activity, underscoring the importance of functional screening.

In the discovery phase, the authors identified antibodies against MHC-I, CD38, CXCR4, and LILRB1 as having strong phagocytosis-activating capabilities in the human system. To validate combinatorial effects, the team tested various dual-antibody combinations and found that anti-CD47 combined with anti-CD38 or anti-LILRB1 significantly enhanced phagocytosis. Subsequently, the authors developed a bispecific antibody platform based on heterodimeric scFv-Fc, using 'knob-in-hole' technology to ensure correct heavy chain pairing. This system enabled rapid cloning, expression, and screening of 156 bispecific antibodies, significantly outperforming traditional IgG-scFv formats with higher expression levels, stronger binding, and reduced RBC binding.

Key Conclusions and Perspectives

  • Functional screening revealed several novel targets (e.g., CD24, CD95, CD40) capable of effectively activating macrophage phagocytosis of A20 cells, suggesting these molecules as potential therapeutic targets to guide future antibody development
  • In the human system, antibodies against CD38, CD71, and LILRB1 exhibited the strongest phagocytic activity when combined with anti-CD47, indicating high translational priority for these targets and supporting their inclusion in bispecific antibody design
  • The bispecific antibody WTa2d1xCD38 demonstrated superior anti-tumor activity in vitro compared to rituximab and daratumumab, and remained highly effective in rituximab-resistant cell lines, suggesting it can overcome CD20 resistance mechanisms and offer a new option for resistant patients
  • WTa2d1xCD38 significantly suppressed Raji cell growth in NSG mouse xenograft models and prolonged survival in an orthotopic brain lymphoma model, demonstrating its activity in complex microenvironments and supporting advancement to preclinical efficacy evaluation
  • WTa2d1xCD38 showed minimal binding to red blood cells, significantly lower than high-affinity SIRPα variants (e.g., CV1) or traditional anti-CD47 antibodies, indicating that low-affinity SIRPα decoys reduce on-target toxicity and improve the therapeutic index, setting a paradigm for safer antibody design

Research Significance and Prospects

This study provides an integrated platform combining high-throughput functional screening with rapid antibody engineering, broadly applicable to other cancers or immune cell types. Particularly for tumors with impaired T-cell function, macrophages become more reliable effectors, and this strategy can accelerate the discovery of innate immune therapies.

Regarding clinical monitoring, the study suggests that tumors lacking MHC-I may be more sensitive to macrophage-based therapies, and future efforts should explore MHC-I expression as a biomarker for patient stratification. Additionally, RBC binding assays should become a routine safety assessment in bispecific antibody development to predict hematological toxicity.

For disease modeling, the study emphasizes the need for immunocompetent or humanized models to more accurately reflect macrophage function. While NSG mice are currently used, their macrophages bind weakly to human SIRPα, potentially underestimating efficacy. Future studies should incorporate humanized immune system mouse models to more accurately evaluate the in vivo activity and safety of bispecific antibodies.

 

 

Conclusion

This study systematically uncovered phagocytosis checkpoints and activating molecules on the surface of B-cell lymphoma through an innovative 'tactical surfaceome analysis' strategy, overcoming the limitations of traditional single-target approaches such as CD20. The authors successfully constructed and screened 156 bispecific antibodies, ultimately identifying WTa2d1xCD38 as the lead candidate, combining high anti-tumor activity with low hematotoxicity. This achievement not only offers new therapeutic hope for patients with B-cell lymphoma, especially those resistant to rituximab, but also establishes a universal platform from functional screening to rapid antibody development. In the future, this strategy could be extended to other hematologic malignancies or solid tumors, accelerating the clinical translation of macrophage-based immunotherapies. More importantly, the study emphasizes the central role of functional validation in target discovery, reminding the scientific community not to rely solely on expression data or binding affinity to predict efficacy. Integrating animal models with in vitro functional assays will accelerate the discovery of next-generation antibody therapeutics, ultimately reshaping the care paradigm for B-cell lymphoma and enabling more precise and safer immunotherapy.

 

Reference:
Carlota Pagès-Geli, Juliano Ribeiro, Thomas Wienclaw, Marta Crespo, and Kipp Weiskopf. High-throughput engineering of bispecific antibodies to enhance macrophage-mediated cytotoxicity of B-cell lymphoma. Nature Communications.
Folding Stability
Prediction of absolute protein stability ΔG by protein sequence inverse folding model ESM-IF. Traditional physical methods (e.g., FoldX, Rosetta, etc.) for predicting protein stability ΔG rely on high-confidence structural pdb, and if there are too many mutations, the structural confidence decreases and the prediction results are poor. Benchmark results at ProteinGym show that the generative model ESM-IF predicts protein mutation stability ΔΔG of DMS data at best-in-class level in zero-shot. The method is an extension of mutation prediction by using the ESM-IF model to directly predict the absolute ΔG value of intact protein folding stability. It was tested with a prediction error RMSE ≈ 1.5 kcal/mol and a correlation coefficient of 0.7, representing a major breakthrough in predicting the folding stability ΔΔG of proteins.