Science explainer

Why intracellular targets need TCR-based T-cell engagers

Antibody-based T-cell engagers can only reach what sits on the surface of a tumor cell. For many solid tumors, the most selective antigens sit inside the cell. T-cell receptor (TCR) based T-cell engagers read those antigens as peptides presented on HLA, and open the intracellular proteome to an off-the-shelf drug.

The surface is a small window

A T-cell engager (TCE) brings a T cell to a tumor cell and triggers killing. An antibody-based TCE does this by binding a protein on the tumor cell surface. That works when the surface protein is abundant on tumor cells and scarce on healthy tissue, which is why the approved TCEs in blood cancers target lineage markers such as CD19, CD20 and BCMA.

Solid tumors are harder. Most surface proteins on a tumor cell are shared with normal tissue, so on-target, off-tumor toxicity limits the dose. Hepatocellular carcinoma (HCC) is a clear example: its most selective antigens sit inside the cell, invisible to conventional antibodies, and the main surface target, GPC3, is still unproven after a decade of attempts [1].

The immune system already looks inside the cell

Every cell continuously breaks down its proteins into short peptides and displays them on HLA molecules at the surface. T cells patrol these peptide-HLA complexes with their T-cell receptors. This is how the immune system detects a virus replicating inside a cell, and how it can detect a cancer-testis antigen or another intracellular protein that a tumor expresses and healthy adult tissue does not.

A TCR-based TCE borrows this mechanism. One arm is a soluble TCR that recognizes a specific peptide-HLA complex on the tumor cell. The other arm engages CD3 on any T cell nearby. The result is a bispecific molecule that redirects T cells to a target an antibody cannot see.

TCR-based T-cell engagers are clinically validated

Tebentafusp is the first and only TCR-based TCE approved in a solid tumor. In metastatic uveal melanoma it improved overall survival (hazard ratio 0.51) even though the response rate was 9%, which shows that T-cell redirection can extend life without the large tumor shrinkage that conventional endpoints look for. No TCR-based TCE has yet reached HCC [2].

Why this is hard to do well

• Finding the TCR. Tumor antigen-specific T cells are rare in blood, fewer than 0.01% of peripheral blood mononuclear cells [3]. Recovering them needs a screen that is both ultra-high throughput and precise.
• Selecting on function, not binding. Native TCRs have modest affinity, and affinity alone does not predict whether a TCR will make a T cell kill a tumor cell. Leads have to be chosen on activation and killing.
• Off-target cross-reactivity. A TCR that recognizes a tumor peptide may also recognize a similar peptide in normal tissue. Clinical investigations have documented such cross-reactivity [4], so safety screening has to cover the whole peptidome, not a panel of proteins.

How ShennonBio approaches it

On TCELERATOR™, patient T cells are paired one to one with antigen-presenting cells in droplets, more than 10 million interactions in a single run [3]. Only the T cells that respond functionally are captured, and their TCRs are sequenced. Because the screen looks for shared antigens presented on common HLA types, the resulting TCRs can be built into an off-the-shelf drug rather than a patient-specific therapy.

At SITC 2025, this approach found 11 validated TCRs to a shared cancer-testis antigen in melanoma patients' T cells. The lead, with no affinity optimization, controlled tumors and extended survival in mice as TCR-T cells [5].

Before any lead advances, SERIS™ scans the candidate TCR against a genome-derived peptidome of more than 10 million peptides to flag potential cross-reactivity with normal tissue. In retrospective testing it ranked clinically validated cross-reactive peptides at the top of the list [1, 4].

This is the path for SBT-425, a TCR-based TCE engaging an intracellular target in hepatocellular carcinoma, with in vivo data expected in the first half of 2027 [1]. See the pipeline for the program overview.

Related reading

• How TCELERATOR screens 10 million single-cell interactions in one run
• Why function beats binding affinity in T-cell engager discovery
• Why a trispecific T-cell engager for small cell lung cancer
• All science explainers

References

1. ShennonBio non-confidential corporate presentation, Nov 2026.

2. Nathan P, et al. Overall survival benefit with tebentafusp in metastatic uveal melanoma. N Engl J Med 2021;385:1196.

3. ShennonBio TCELERATOR technical presentation, Oct 2026 (internal validation data; available under a confidentiality agreement).

4. Linette GP, et al. Cardiovascular toxicity and titin cross-reactivity of affinity-enhanced T cells in myeloma and melanoma. Blood 2013;122:863.

5. Li K, et al. J Immunother Cancer 2025;13(Suppl 2):A383 (SITC 2025).