Company News

News Categories

Hybridoma Versus Display Technologies in Modern Monoclona...

Smarter Cancer Targeting: How Bispecific Antibodies Are R...

Designing Linker Sequences and Their Impact on scFv Stabi...

Downstream Purification Strategies for High-Quality VHH A...

Cancer-Driven Somatic Mutations Enrich Alzheimer's Diseas...

Complement C5: Mechanisms, Functions, and FDA-Approved Th...

Fine-Tuning SELEX Stringency Conditions to Improve Aptame...

Antigen Preparation Strategies to Maximize Immune Respons...

Designing Focused Peptide Libraries Based on Structural a...

Principles and Applications of Surface Plasmon Resonance ...

recommended service

Smarter Cancer Targeting: How Bispecific Antibodies Are Rewriting the Rules of Oncology

2026-08-14

IntroductionNEWS

For decades, cancer researchers faced a major frustration. The signaling pathways that drive tumor growth are also required for healthy tissue function. Blocking pathways like Wnt or EGFR in a cancer cell sounds simple. However, the same inhibition destroys the intestinal lining, fractures bones, or blisters skin. A new study from researchers at Genentech published in Science Advances (Kschonsak et al., 2026) offers a compelling answer. The strategy is built on a simple idea. They paired a weak inhibitor of an oncogenic receptor with a high-affinity arm that targets a tumor-enriched surface protein. Because of this, the drug activates only where it is needed.

The Core Strategy: Guided, Conditional InhibitionNEWS

The conceptual backbone of this work rests on two insights. First, pathway-associated toxicities tend to be tissue-specific. Wnt inhibitors harm the gut and bone. EGFR inhibitors damage the skin, and FGFR1 inhibitors disrupt kidney phosphate regulation. This specificity implies that the safe zone is much larger than traditional antibody design assumed. Second, single-cell RNA sequencing (scRNA-seq) provides the resolution to find which surface proteins are expressed on tumor cells. Crucially, these proteins must be missing from the specific cell types driving toxicity, which are usually stem cell populations.

The team exploited these two facts. They engineered bispecific antibodies (bsAbs) where one arm binds weakly to the therapeutic target, such as Frizzled/FZD, EGFR, or FGFR1. Affinities were deliberately attenuated above 50 nM. This makes the arm too weak to work by itself. The second arm binds with high affinity to a tumor-enriched anchor receptor, like TROP2 or CEACAM6. On tumor cells expressing the anchor, local concentration of the bsAb occurs. Avidity-driven co-engagement of both receptors restores potent inhibitory activity. On normal cells that lack the anchor, the weak inhibitory arm never gains enough traction to cause harm. The knobs-into-holes IgG1 format was used with effector-function-ablating mutations. This framework was applied across 3 oncogenic targets. Thus, the group established it as a broadly transferable platform rather than a one-off solution.

Send Inquiry

Contact Us For Best Would you Like to Know more We can Give you the answer, For inquiries about our products and services. please leave your e-mail to us and will reply within 24 hours.

Click for inquiry

Core ResultsNEWS

Wnt Inhibition, Unavoidable Toxicity

Fig 1 Targeting FZD in RNF43 mutant PDAC is limited by toxicity

The study opens by validating the clinical problem. DepMap functional screen data reveal that FZD5 is a lineage-selective dependency in pancreatic ductal adenocarcinoma (PDAC). Meanwhile, genes like CTNNB1 and TCF7L2 dominate colon adenocarcinoma. In the RNF43-mutant HPAF-II cell line, clonal growth was halted by doxycycline-induced FZD5 knockout. This worked as effectively as the porcupine inhibitor LGK974. Transcriptional markers confirmed on-target Wnt pathway suppression, showing a drop in AXIN2 and NKD1, and a rise in MUC5AC. Critically, a panel of PDAC lines was tested. The pan-FZD inhibitor F2.I outperformed the FZD5/8-selective antibody. This suggests that multi-receptor blockade is required for a durable response. Yet, mice received just 2 mg/kg of F2.I or 5 mg/kg of the FZD5/8 antibody. Within 8 days, rapid body weight loss and severe villous atrophy of the small intestine followed. The figure makes the therapeutic tension vivid. Efficacy and toxicity arise from the exact same molecular event. Simply titrating the dose is not a workable solution.

Engineering Selective FZD Blockade

Fig 2 Design and generation of a bispecific tumor targeting anti-FZD

The team developed a solution pipeline to fix the toxicity issue. They engineered an attenuated pan-FZD variant called F2.Iv2. This variant has a binding affinity >50 nM. On normal colon organoids, F2.Iv2 reduced activity by more than 100-fold compared to parental F2.I. Mice tolerated 90 mg/kg of F2.Iv2 with only minor intestinal findings. By contrast, a tiny 2 mg/kg dose of F2.I caused profound weight loss. To find good anchor receptors, the researchers cross-referenced bulk RNA-seq data from 36 human PDAC tumors against scRNA-seq datasets of human intestine and bone marrow. They looked for receptors highly expressed in PDAC tumor epithelium but completely missing from LGR5⁺ intestinal stem cells and LEPR⁺ bone marrow stem cells. TROP2 (encoded by TACSTD2) and CEACAM6 came out as the top candidates. Literature using immunohistochemistry confirmed these receptors are not in gut and bone compartments. The researchers paired these anchors with F2.Iv2 to make bispecific antibodies (bsAbs). Both TROP2- and CEACAM6-targeted bsAbs restored sub-nanomolar growth inhibition in HPAF-II cells and patient-derived PDAC organoids. Meanwhile, human colon organoids do not have these anchor receptors, so they remained fully viable. qPCR checked the mechanism. It confirmed canonical Wnt pathway suppression via drops in AXIN2, NKD1, and MKi67, and an increase in MUC5AC. This proved the mechanism was on-target rather than random cytotoxicity.

Efficacy Without Intestinal Damage

Fig 3 Tumor-targeted anti-FZDs are efficacious in in vivo tumor model with minimal intestinal toxicity

Animal models confirmed that the tumor-targeting concept works in vivo. In HPAF-II xenografts, researchers gave a single 30 mg/kg dose of aTROP2/F2.Iv2. Within 48 hours, this dose robustly suppressed AXIN2, NKD1, and LGR5 inside tumor tissue. This pharmacodynamic data shows that the bsAb actually reaches its target and shuts down the Wnt pathway selectively inside the tumor microenvironment. Efficacy studies used doses of 5 and 30 mg/kg. Both doses caused substantial tumor volume reductions. Crucially, mouse body weights remained stable. This is a huge contrast to the rapid deterioration caused by F2.I. A mouse-cross-reactive antibody variant, aTROP2.mAb1/F2.Iv2, cleared from the serum faster than a non-targeting control. This faster clearance happens because the antibody distributes into normal tissues that express TROP2. Even so, the drug maintained serum concentrations roughly 200-fold above the in vitro IC₅₀ across the entire dosing interval, which is more than enough for antitumor activity. For safety testing, immunocompetent non-tumor-bearing C57BL/6 mice got three doses of 30 mg/kg. These mice showed 0% body weight loss and had histologically normal duodenum. This was completely different from the severe intestinal destruction caused by just a single 2 mg/kg dose of the unguided F2.I inhibitor.

Internalization Drives Potency

Fig 4 Tumor-targeted Abs can turn almost inert FZD binders into potent blockers

A mechanistic study showed that avidity restoration does not fully explain the potency. The team replaced F2.Iv2 with FZD.mAb1, which is an independent anti-FZD antibody. On its own, FZD.mAb1 has almost zero growth-inhibitory activity. However, the resulting TROP2- and CEACAM6-targeted bsAbs still killed cells effectively. Their IC50 values were only slightly higher than the F2.Iv2-based molecules. A pH-sensitive fluorophore internalization assay was performed. The data showed that bsAbs containing either anti-FZD arm internalized faster than the TROP2 antibody by itself. This indicates that co-crosslinking TROP2 and FZD accelerates endocytic uptake of both receptors at the same time. This forced co-internalization traps FZD inside the cell. Because it is sequestered, FZD cannot bind Wnt ligands at the cell membrane. This mechanism works even if the anti-FZD arm has very weak intrinsic blocking potency. Bulk RNA-seq data confirmed the outcome. Both F2.Iv2- and FZD.mAb1-based bsAbs replicated the transcriptional signature of direct Wnt inhibition. The signatures correlated at r=0.767 and were confirmed by Hallmark GSEA. Standalone FZD.mAb1 did not cause any meaningful Wnt gene-set suppression.

Proof of Concept Broadens

Fig 5 Potent tumor-targeted antagonists of EGFR and FGFR1 that spare cell types responsible for toxicity

The last figure shows the platform can expand to EGFR and FGFR1. For FGFR1, scRNA-seq data from human kidney showed that FGFR1 and TACSTD2 are expressed in totally different renal cell types, specifically proximal tubules versus ductal cells. This spatial separation allows selective targeting. A TROP2/aFGFR1 bsAb completely blocked FGF1-induced FGFR1 and PLCγ phosphorylation in TROP2-engineered breast cancer cells. At the same time, FGFR1 signaling in the G-402 kidney cell line was left completely intact. For EGFR, skin scRNA-seq data showed that CEACAM6 is nearly absent in keratinocytes while TACSTD2 is highly expressed. Therefore, CEACAM6 was chosen as the better anchor receptor. The CEACAM6/aEGFR bsAb blocked the EGFR-dependent CRC cell line NCI-H508. It was just as potent as bivalent cetuximab. However, it required more than a 100-fold higher concentration to cause any effect on human adult keratinocytes. This represents a major improvement in the therapeutic window compared to normal EGFR blockade.

Implications for Next-Generation Targeted TherapyNEWS

This work redefines how we target receptors in oncology. A target does not need tumor-exclusive expression. Very few receptors can meet that standard anyway. Instead, the framework only requires that the anchor receptor is missing from the specific cell type that causes dose-limiting toxicity. Single-cell transcriptomics allows researchers to find these differences at scale. This converts previously dangerous targets into safe, druggable opportunities. As platforms like this move toward clinical trials, they provide real options for patients whose treatments were previously stopped not by poor efficacy, but by a lack of safety margin.

This study establishes a paradigm-shifting platform for oncology by demonstrating that bispecific antibodies can convert previously undruggable targets with on-target toxicity into safe therapeutic opportunities through tumor-enriched anchor receptors, effectively decoupling efficacy from dose-limiting adverse effects. Alpha Lifetech offers specialized CRO services in Phage Display-based Antibody Discovery - including VHH, Fab, scFv, Aptamer, and Peptide development - to help you translate cutting-edge genomic insights into precise biologic therapeutics.

FAQsNEWS

1. Why have traditional cancer pathway inhibitors struggled with safety, and what makes this approach different?

Most oncogenic pathways that cancer cells exploit for uncontrolled growth are the same pathways that healthy tissues depend on for routine maintenance and repair. When a drug blocks the Wnt pathway to slow a pancreatic tumor, it also disrupts the intestinal stem cells that regenerate the gut lining every few days. The result is a narrow or even nonexistent window between an effective dose and a harmful one.

What separates the bispecific antibody strategy described in this research is a shift in targeting logic. Instead of asking whether a receptor is expressed exclusively on tumors, the researchers asked a more practical question: is the anchor receptor absent from the specific cell population responsible for toxicity? That is a much easier bar to clear. By pairing a deliberately weakened inhibitory arm with a high-affinity arm directed at a tumor-enriched surface protein, the drug essentially becomes context-dependent. It gains full potency only on cells that carry both the anchor receptor and the therapeutic target, leaving cells that express only one of the two largely unaffected.
2. How did the researchers identify TROP2 and CEACAM6 as suitable anchor receptors for Wnt pathway targeting?

The selection process combined two levels of data analysis to arrive at candidates that were both tumor-relevant and safe from a toxicity standpoint.

(i) Bulk RNA sequencing of 36 human pancreatic ductal adenocarcinoma tumors and matched normal pancreatic tissue was used first to generate a broad list of surface receptors that were significantly upregulated in tumor epithelium.
(ii) Single-cell RNA sequencing atlases of the human intestine and bone marrow were then layered on top of that list. The goal was to filter out any receptor expressed in LGR5-positive intestinal stem cells or LEPR-positive bone marrow stem cells, since those are the cell populations that Wnt inhibitors are known to damage.
(iii) TROP2 and CEACAM6 cleared both filters. Neither receptor showed meaningful expression in those critical stem cell compartments, and published immunohistochemistry data independently confirmed the absence of TROP2 protein in intestinal and bone tissue.

This two-step funnel, moving from bulk expression data to single-cell resolution, is one of the more transferable methodological contributions of the work. It provides a reproducible framework for identifying anchor receptors in other disease contexts.
3. What is actually happening at the molecular level when these bispecific antibodies inhibit tumor cells?

The mechanism turns out to involve more than simple avidity restoration, which was the team's original hypothesis. Two processes appear to work together.

(i) The anchor arm binds tightly to TROP2 or CEACAM6 on the tumor cell surface, concentrating the bispecific antibody locally and increasing the effective on-cell concentration of the weakened inhibitory arm. This allows the inhibitory arm to engage its target receptor despite its low standalone affinity.
(ii) Once bound to both receptors simultaneously, the bispecific antibody triggers accelerated co-internalization. Live cell imaging showed that bispecific antibodies were taken up faster than the TROP2 antibody alone, pulling the therapeutic target receptor into the endosomal compartment along with the anchor receptor. This physically removes the target from the cell surface, preventing the cancer cell from receiving the growth signal regardless of how much ligand is present in the surrounding environment.
(iii) Importantly, transcriptomic analysis confirmed that bispecific antibody treatment produces a gene expression signature nearly identical to that of a direct, high-potency inhibitor, validating that the downstream biological effect is genuine pathway suppression rather than an off-target artifact.
4. Does this strategy work beyond Wnt signaling, or is it specific to pancreatic cancer?

The researchers deliberately tested the platform across three distinct oncogenic targets to demonstrate generalizability, and the results across all three were encouraging.

(i) For FGFR1 in breast cancer, single-cell kidney data confirmed that TROP2 and FGFR1 are expressed in entirely separate renal cell types. A TROP2-targeted FGFR1 bispecific antibody fully suppressed growth factor signaling in TROP2-positive breast cancer cells while leaving kidney cell signaling completely intact, addressing the hyperphosphatemia risk that has historically limited FGFR inhibitor development.
(ii) For EGFR in colorectal cancer, skin transcriptomic data revealed that CEACAM6 is nearly absent from keratinocytes while being well expressed on colorectal tumor cells. A CEACAM6-targeted EGFR bispecific antibody matched the potency of cetuximab against colorectal cancer cells but required concentrations more than 100 times higher to produce any effect on adult skin keratinocytes, dramatically improving the predicted therapeutic window for EGFR blockade.

The consistent performance across three targets with distinct biology and distinct toxicity profiles suggests the platform reflects a generalizable principle rather than a coincidence specific to one receptor pair.
5. What would need to happen before a therapy like this reaches patients?

The current study establishes preclinical proof of concept, and the authors are transparent about the additional work required before clinical translation becomes feasible.

(i) Species cross-reactivity must be confirmed. Antibody arms need to bind equivalently to mouse, cynomolgus monkey, and human versions of both target proteins so that animal studies can accurately predict on-tumor activity and off-tumor safety at the same time.
(ii) Formal dose-ranging studies linking drug exposure to tumor response and to any emerging toxicity signals will be necessary to define a viable therapeutic index in the clinic.
(iii) Long-term, multi-dose safety studies with thorough histological analysis across multiple organ systems are required to characterize the full safety profile, including any unexpected effects introduced by the anchor receptor itself.
(iv) Non-human antibody sequences, such as the rabbit-derived TROP2 antibody used in some experiments, will need to be humanized to reduce the risk of immunogenic reactions in patients.
(v) Manufacturability factors including antibody stability, aggregation tendency, and viscosity at high concentrations will need to meet the standards required for a commercial biopharmaceutical product.

Each of these steps is well within the scope of established drug development practice. The scientific framework itself is now validated; the remaining work is a matter of careful, systematic execution.

ReferenceNEWS

[1] Yvonne T. Kschonsak et al. ,Tumor-targeted bispecific antibodies effectively inhibit oncogenic pathways while minimizing toxicity.Sci. Adv.12,eadx3959(2026).DOI:10.1126/sciadv.adx3959
[2] Shyam Sunder S, Sharma UC, Pokharel S. Adverse effects of tyrosine kinase inhibitors in cancer therapy: pathophysiology, mechanisms and clinical management. Signal Transduct Target Ther. 2023 Jul 7;8(1):262. doi: 10.1038/s41392-023-01469-6. PMID: 37414756; PMCID: PMC10326056.
[3] Dumontet C, Reichert JM, Senter PD, Lambert JM, Beck A. Antibody-drug conjugates come of age in oncology. Nat Rev Drug Discov. 2023 Aug;22(8):641-661. doi: 10.1038/s41573-023-00709-2. Epub 2023 Jun 12. PMID: 37308581.