The central achievement of the study is a bispecific TCR-like antibody, designated 2F7, which recognizes a fragment of the PRAME protein displayed on the cell surface by the HLA-A*24:02 molecule. To appreciate why this matters, it helps to understand how the immune system inspects cells for danger. Inside virtually every nucleated cell, proteins are constantly being chopped into short peptide fragments by the cellular machinery. A sample of these fragments is loaded onto major histocompatibility complex class I molecules and transported to the cell surface, where they are displayed to patrolling cytotoxic T lymphocytes. In a healthy cell, the peptides are ordinary self-proteins and are ignored. In a cancerous or infected cell, peptides derived from abnormal proteins betray the cell’s true identity, triggering a lethal immune attack.

The catch is that T cells cannot see inside a cell. They only recognize peptides when they are presented in the groove of an HLA molecule, and they do so using highly specialized receptors known as T cell receptors, or TCRs. For decades, immunologists have dreamed of harnessing this recognition system with antibodies — molecules that can be engineered with exquisite specificity, manufactured at scale, and deployed as drugs. This is the rationale behind TCR-like antibodies: artificial binding proteins generated in the laboratory that mimic the specificity of a natural T cell receptor by recognizing a specific peptide-MHC complex. Because they can distinguish a single peptide antigen presented on an HLA scaffold, TCR-like antibodies effectively give therapeutic agents X-ray vision into the interior of a cell, allowing them to target proteins that were previously considered undruggable precisely because they are buried inside the tumor cell and never appear on the surface in conventional form.

Yet almost all TCR-like antibodies developed to date have targeted peptides presented by HLA-A02:01, better known as HLA-A2. This allele is indeed the most common MHC class I variant in the world, but its dominance in the scientific literature reflects convenience as much as biology. HLA genes are among the most polymorphic in the human genome, and peptide presentation is strictly allele-dependent: a given peptide will only be displayed if the individual carries a compatible HLA molecule. A TCR-like antibody built against HLA-A2 presented peptides is therefore useless in patients who do not carry that allele. With HLA-A2 present in only about half of people of European descent and substantially lower proportions in many other populations, entire regions of the world have effectively been locked out of this therapeutic strategy. East Asian populations in particular carry a different dominant allele: HLA-A24:02, the second most frequent MHC class I allele worldwide and the single most common one across much of Asia.

The Korean team chose their target accordingly, and their choice of antigen was equally deliberate. PRAME — preferentially expressed antigen in melanoma — belongs to the cancer-testis antigen family, a class of proteins whose expression is normally restricted to the male germline but which become aberrantly reactivated in a wide spectrum of malignancies. Because healthy adult tissues outside the testes do not produce PRAME, and because the testes are an immunologically privileged site that does not present class I peptides to the immune system in the usual way, a therapy targeting PRAME-derived peptides carries an inherently favorable safety profile. PRAME is expressed in melanoma, lung cancer, breast cancer, leukemia, head and neck cancers, and numerous other tumor types, making it one of the most broadly applicable tumor-associated antigens known. The specific peptide targeted in this study, PRAME301–309, is a nine-amino-acid fragment of the protein presented by HLA-A*24:02.

Finding an antibody that can recognize such a structure is no small feat. Peptide-HLA complexes are notoriously difficult targets: the HLA molecule itself is present on every cell in the body, so the antibody must bind in a way that contacts the specific peptide nestled in the HLA groove while tolerating or ignoring the surrounding HLA framework. The researchers accomplished this using phage display, a powerful in vitro evolution technique in which billions of antibody fragments are displayed on the surface of bacteriophages — viruses that infect bacteria — and screened iteratively for the rare clones that bind the desired target. From this molecular haystack, the team isolated a single-chain variable fragment, or scFv, designated 2F7, specific for the PRAME301–309/HLA-A*24:02 complex.

An scFv, however, is only a building block. To convert it into a therapeutic molecule, the researchers reformatted 2F7 into a bispecific T-cell–engaging antibody, commonly abbreviated as a T-cell engager or bsTLA in this context. Bispecific antibodies are engineered proteins with two different binding arms: one arm recognizes the tumor-associated target — in this case, the PRAME301–309/HLA-A24 complex — while the other arm binds CD3, a signaling component of the T cell receptor complex expressed on essentially all T cells. The result is a molecular matchmaker. By physically bridging a cytotoxic T cell to a cancer cell displaying the target peptide-HLA complex, the bispecific antibody forces the immune cell into intimate contact with its prey, triggering activation, release of cytotoxic granules containing perforin and granzymes, and the destruction of the targeted tumor cell. This mechanism effectively commandeers any T cell in the patient’s body, regardless of the natural specificity of its T cell receptor, and redirects it against the tumor.

The team then put 2F7 through its paces in the laboratory. Using human peripheral blood mononuclear cells — PBMCs, the mixed population of immune cells found in circulating blood — together with cancer cell lines engineered or selected to present the PRAME301–309/HLA-A24 complex, the researchers demonstrated two critical properties. First, the 2F7 bispecific antibody bound specifically and selectively to cells displaying the target peptide-HLA complex, confirming that the phage display campaign had genuinely yielded a TCR-like specificity rather than an antibody that merely recognized HLA-A24 regardless of its peptide cargo. This distinction is paramount: an antibody that bound all HLA-A24 molecules indiscriminately would attack every cell in an HLA-A24-positive patient, with catastrophic consequences. Second, and more importantly, the antibody successfully redirected PBMC-derived T cells to kill target cells presenting the peptide-HLA complex, providing proof of concept that the molecule can orchestrate tumor cell killing through endogenous T cells.

The implications of the work extend well beyond a single antibody. Roughly speaking, a therapy built on HLA-A*24:02 could be applicable to the substantial majority of patients in East Asian countries such as Korea, Japan, and China, as well as sizable patient populations elsewhere — a demographic reach that HLA-A2-based approaches simply cannot match. For South Korea, where much of the biotechnology and cell therapy industry is concentrated, the development of an allele-appropriate TCR-like antibody represents a meaningful step toward immunotherapies designed for the patients most likely to use them. It also establishes a template: the same pipeline of phage display selection, peptide-HLA specificity validation, and bispecific reformatting can in principle be applied to other cancer-testis antigens and other HLA alleles, gradually closing the coverage gaps left by an HLA-A2-centric field.

The researchers emphasize that the current study is an in vitro characterization — the experiments were performed with cultured cells and donor-derived immune cells, not in patients. The path from laboratory validation to clinical application involves additional hurdles: confirmation of efficacy and safety in animal models, manufacturing development, regulatory review, and ultimately clinical trials in which questions of dosing, cytokine toxicity, on-target off-tumor effects, and tumor immune evasion will need to be answered. Cytokine release syndrome, the systemic inflammatory reaction that has complicated the clinical use of other T-cell engagers such as blinatumomab, remains a consideration for any molecule that activates T cells systemically. The authors also note that a patent application pertaining to the antibodies reported in the work is in preparation, a familiar marker of translational intent.

Still, the significance of the achievement is hard to overstate. The T-cell engager concept has already proven clinically transformative in hematologic malignancies, but solid tumors and broader patient populations have remained stubbornly difficult targets. TCR-like antibodies such as 2F7 offer a way to bring the precision of T-cell recognition — the ability to see intracellular antigens through the window of peptide-HLA presentation — into an antibody format that can be produced, standardized, and administered off the shelf, without the complexity and cost of engineering a patient’s own cells. By demonstrating that such a molecule can be built against HLA-A*24:02, the most common class I allele in East Asia and the second most common worldwide, the Ewha Womans University team and their collaborators have broadened the addressable patient population for this technology and issued a quiet challenge to the field: immunotherapy for the world cannot be built on one HLA allele alone. As PRAME-expressing tumors account for a substantial share of human cancers, and as HLA-A24 carriers number in the hundreds of millions, the 2F7 antibody may mark the beginning of a more geographically and genetically inclusive era of targeted cancer immunotherapy.

Historically, the molecular landscape of cancer has posed enormous challenges, particularly in AML. This aggressive blood cancer exhibits a complex network of pathways essential not only to malignant cells but also to normal hematopoietic stem cells, thus creating a precarious therapeutic balance. Conventional drugs, albeit somewhat effective, often falter due to their inability to distinguish thoroughly between malignant and normal cells, resulting in substantial toxicity and poor patient tolerance. This scientific impasse has sustained an urgent demand for selective molecular targets—biomarkers that are expressed predominantly or exclusively on cancerous cells.

The Flynn group’s discovery capitalizes on an unusual feature: the ectopic localization of the RNA-binding protein NPM1 to the exterior of AML cells. While NPM1 traditionally functions within the nucleolus as a chaperone for ribosomal biogenesis and genomic stability, its aberrant expression on the cell surface of cancer cells marks a profound departure from its canonical role. Detailed investigations revealed that cell-surface NPM1 is dramatically upregulated in leukemic cells, with expression levels exceeding those found on healthy blood stem cells by over 100-fold. This significant differential creates a therapeutically exploitable target that, until now, remained concealed within the interior of the cell.

The team elucidated the mechanistic underpinnings of this phenomenon in the context of glycoRNAs—an emerging class of glycoconjugated RNA molecules residing on the cell exterior, which form organized clusters with RNA-binding proteins including NPM1. Prior foundational work has characterized these glycoRNA-protein complexes as novel signaling platforms modulating cellular communication with the microenvironment. This groundbreaking concept redefines the understanding of cell-surface biology, highlighting an uncharted molecular landscape ripe for targeted intervention.

Leveraging this insight, Flynn and colleagues engineered monoclonal antibodies specifically directed against NPM1 presented on the surface of AML cells. These antibodies demonstrated potent anti-leukemic efficacy across multiple preclinical in vivo models, selectively eliminating malignant cells while sparing normal hematopoietic populations. Such specificity is crucial as it addresses one of the most stubborn obstacles in AML treatment—the preservation of healthy bone marrow function during therapy. Notably, the antibodies also effectively targeted leukemic stem cells, the elusive subpopulation responsible for disease initiation, persistence, and relapse.

The impact of targeting leukemic stem cells cannot be overstated. These cells exhibit remarkable resistance to conventional chemotherapies and are often responsible for the clinical recurrence of AML. By attacking these cells head-on through a uniquely surfaced antigen like NPM1, the therapeutic paradigm shifts from merely controlling disease to potentially achieving durable remission or cure. In murine models, this strategy extended survival and markedly reduced disease burden, with no observed off-target toxicity, emphasizing the treatment’s clinical promise.

Beyond leukemia, the research explored the broader oncological relevance of cell-surface NPM1. Screening an extensive panel of 47 human and murine solid tumor models unveiled variable but significant expression of cell-surface NPM1 across many tumor types, including prostate and colorectal carcinomas. These findings suggest a wider applicability of NPM1-targeting antibodies, potentially expanding immunotherapy’s arsenal against notoriously treatment-resistant solid tumors.

The identification of NPM1 as a cell-surface antigen in solid tumors is particularly compelling given the historical difficulty of finding cancer-selective surface markers for these malignancies. Cancers like colorectal carcinoma have long evaded effective immune targeting due to the scarcity of unique markers distinguishable from normal tissue. The cell-surface presentation of NPM1 thus represents a potential ‘molecular handle’ for immune system engagement, a prospect that could reinvigorate therapeutic strategies for multiple cancers.

Crucially, the research underscores the newly appreciated biology of glycoRNAs and RNA-binding proteins as a rich source of tumor-associated antigens. The clustering of these molecules on the cell surface appears not to be a random occurrence but an orchestrated phenomenon potentially advantageous to tumor survival and immune evasion. The team’s future investigations aim to decode the biological imperatives underpinning the externalization of NPM1 and to identify additional molecular candidates within these clusters that could serve as targets or biomarkers.

The discovery that malignant cells co-opt an RNA-binding protein, traditionally intracellular, and mobilize it to the cell membrane hints at a novel tumor strategy that may confer advantages such as altered signaling, adhesion, or immune modulation. Understanding these dynamics will be critical to refining antibody-based therapeutics and possibly integrating them with other modalities, including cellular therapies and immune checkpoint inhibitors.

To translate these foundational findings into clinical impact, Boston Children’s Hospital has already pursued intellectual property protections domestically and internationally. This strategic move paves the way for the development of antibody therapies targeting NPM1, with the potential to enter early-phase clinical trials and ultimately offer new hope to patients with aggressive hematologic and solid malignancies.

The collaboration among interdisciplinary teams spanning molecular biology, oncology, immunotherapy, and structural biochemistry highlights the power of cross-sector partnerships in unearthing novel therapeutic targets. The convergence of expertise in glycoRNA biology, stem cell research, and antibody engineering illustrates a modern scientific approach to solving intractable problems in medicine.

In summary, Dr. Ryan Flynn’s team has illuminated a captivating facet of cancer biology—the aberrant cell-surface expression of an RNA-binding protein—and harnessed it into an actionable therapeutic target. By shifting the paradigm toward precision targeting of cancer stem cells with minimal collateral damage, their work charts a course for next-generation cancer therapies. As future studies delve deeper into the mechanisms and clinical translation, this discovery holds transformative potential for millions battling AML and other formidable cancers, marking a true milestone in the quest for safer, more effective treatments.

Subject of Research: Treatment of acute myeloid leukemia and solid tumors through targeting cell-surface RNA-binding proteins, specifically NPM1.

Article Title: Treatment of acute myeloid leukemia models by targeting a cell-surface RNA-binding protein

News Publication Date: 23-Apr-2025

Web References:
DOI: 10.1038/s41587-025-02648-2
Flynn Lab at Boston Children’s Hospital
Cambridge Stem Cell Institute

Keywords:
Cancer stem cells, RNA binding proteins, Myeloid leukemia, Gene targeting, Molecular targets, Stem cell therapy, Antibody therapy, Monoclonal antibodies, Cell surface receptors