TL;DR
Researchers have developed NXA01, a novel antibody that targets CD24, a "don't-eat-me" signal used by cancer cells to evade the immune system. Unlike previous attempts, NXA01 is highly selective, specifically targeting the "stripped-down" (hypoglycosylated) version of CD24 found on tumors while sparing healthy cells, potentially offering a safer and more potent immunotherapy.
What Problem Does This Paper Address?
The immune system is naturally equipped to clear abnormal cells, but cancers often deploy "molecular camouflage" to stay hidden. One such shield is CD24, a protein overexpressed in various cancers (breast, ovarian, lymphoma). CD24 binds to Siglec-10 on macrophages, sending a "don't-eat-me" signal that prevents the immune system from engulfing the tumor.
While targeting CD24 is a promising strategy, there is a major catch: CD24 is also present on some healthy cells, such as B cells and granulocytes. A non-selective antibody would attack these healthy cells, leading to "on-target, off-tumor" toxicity. The challenge addressed here is creating an antibody that can distinguish between "cancerous" CD24 and "healthy" CD24.
Methods
The team at NeoX Biotech used a multi-pronged approach to discover and refine NXA01:
- Differential Immunization: They immunized mice using human CD24 purified from two different sources: E. coli (which produces non-glycosylated proteins) and 293T mammalian cells (which produce fully glycosylated proteins). This allowed them to screen for antibodies that recognize specific structural states of the protein.
- Precision Screening: They used Surface Plasmon Resonance (SPR) to measure binding affinity and Flow Cytometry (FACS) to ensure the antibody bound to tumor cells but ignored normal human blood cells.
- Protein Engineering: The lead candidate was humanized in silico (using computer modeling) and further optimized through affinity maturation via saturation mutation to ensure it binds its target as tightly as possible.
- Mechanistic Assays: They performed co-culture studies to measure ADCC (Antibody-Dependent Cellular Cytotoxicity) and ADCP (Antibody-Dependent Cellular Phagocytosis)—essentially testing if the antibody successfully "tags" cancer cells for destruction by Natural Killer cells and macrophages.
Datasets
The researchers evaluated NXA01 across several biological models:
- Cancer Cell Lines: NALM6 (B-cell precursor leukemia) and MDA-MB-468 (triple-negative breast cancer).
- Normal Primary Cells: Human granulocytes and B cells (used as a "negative control" to test selectivity).
- In Vivo Models: MC38 cells (mouse colon cancer) engineered to express human CD24, along with NALM6 and MDA-MB-468 xenografts in mice.
Key Insights and Ideas
- The Glycan "Tell": The most significant finding is that NXA01’s selectivity is driven by the glycosylation state of CD24. Cancer cells often have "messy" or incomplete sugar coatings (hypoglycosylation). NXA01 preferentially binds to these "naked" versions of the protein. When the researchers enzymatically stripped the sugars off healthy cells, NXA01 suddenly started binding to them, proving that the sugar coating is what normally protects healthy cells from this drug.
- Dual-Action Weaponry: NXA01 doesn't just block the "don't-eat-me" signal (the checkpoint blockade). It also actively recruits immune cells to kill the tumor via ADCC and ADCP. This "one-two punch" makes it more effective than a simple blocking agent.
- Synergy with Existing Drugs: The study found that NXA01 enhanced the phagocytic activity of tafasitamab (an anti-CD19 antibody), suggesting it could be a powerful partner in combination therapies.
Key Results
- High Affinity: NXA01 binds to CD24 with a very strong affinity of 1.5 nM.
- Potent Blockade: It successfully blocked the CD24-Siglec10 interaction with an EC50 of 8 nM.
- Tumor Selectivity: In mixed cultures of tumor and normal cells, NXA01 mediated the killing of cancer cells (EC50 < 10 nM) while leaving normal B cells untouched.
- In Vivo Efficacy: In mouse models, NXA01 treatment led to dramatic tumor growth inhibition:
- 104% inhibition in MC38-hCD24 models (indicating tumor regression).
- 105% inhibition in NALM6 models.
- 51% inhibition in MDA-MB-468 breast cancer models.
Limitations and Future Work
While the results are highly promising, the paper notes a few areas for further development:
- Variable Efficacy: The inhibition in the breast cancer model (51%) was notably lower than in the leukemia/lymphoma models (100%+). This suggests that tumor microenvironments or varying CD24 glycan profiles in solid tumors may impact performance.
- Clinical Translation: As this is a poster abstract, the data is pre-clinical. The authors state that IND-enabling studies (the final safety and manufacturing tests required before human trials) are currently ongoing. The true test of its "best-in-class" potential will be its safety profile in human patients.
Summary updated