Dual-ligand engineering of NK cell surfaces using lipid conjugates: Harnessing bispecific targeting to CD44 and sialic acid for solid tumor treatment

초록

CD44, highly expressed on cancer cell surfaces, has frequently been targeted by hyaluronic acid (HA)-functionalized cell and drug delivery strategies for targeted therapy. However, CD44-mediated single targeting has limitations: (1) CD44 expression varies across tumor types compromising consistency and efficacy; and (2) CD44 is also expressed on normal cells potentially decreasing tumor specificity. To overcome these limitations, we selected sialic acid (SA) as a complementary target which is overexpressed in various cancers and specifically recognized by phenylboronic acid (PBA). We designed a lipid-poly(ethylene glycol) (PEG)-HA-PBA conjugate that could stably anchor on natural killer (NK) cell membranes via hydrophobic interactions. Co-presentation of PBA and HA on the cell surface was intended to broaden targeting capability by enabling simultaneous engagement of CD44 and SA reducing off-target interactions. This dual-targeting system was applied to pancreatic and colorectal cancer models which show heterogeneity of CD44 and SA expression profiles. Compared to single-ligand targeting, dual-targetable engineered NK cells exhibited enhanced cancer recognition and cytotoxic activity in vitro. These findings were further corroborated by improved tumor growth suppression and the inhibition of metastatic progression in vivo. Notably, the treatment led to significant tumor regression in a colorectal cancer xenograft model. Furthermore, safety assessments confirmed that the engineered NK cells maintained high tumor specificity without off-target cytotoxicity against normal fibroblasts and major organs. Collectively, this dual-ligand surface engineering strategy addresses key limitations of single-ligand targeting approaches. It demonstrates the potential of a modular NK cell-based platform for multi-targeted cancer immunotherapy under physiologically relevant conditions. © 2026 Elsevier B.V.

키워드