The manipulation of cell–cell interactions promotes the study of multicellular behavior, but it remains a great challenge for programming multicellular assembly in complex reaction pathways with multiple cell types. Here we report a DNA reaction circuit-based approach to cell–surface engineering for the programmable regulation of multiple cell–cell interactions. The DNA circuits are designed on the basis of a stem-loop-integrated DNA hairpin motif, which has the capability of programming diverse molecular self-assembly and disassembly pathways by sequential allosteric activation. Modifying the cell surface with such DNA reaction circuits allows for performing programmable chemical functions on cell membranes and the control of multicellular self-assembly with selectivity. We demonstrate the selective control of targeting the capability of natural killer (NK) cells to two types of tumor cells, which show selectively enhanced cell-specific adaptive immunotherapy efficacy. We hope that our method provides new ideas for the programmable control of multiple cell–cell interactions in complex reaction pathways and potentially promotes the development of cell immunotherapy.

Accurate profiling of metabolites in biofluids provides a feasible approach to disease diagnosis and therapeutic evaluation in clinics. However, current material‐based analytical platforms for in vitro metabolite detection often requires tedious sample pretreatment. To address this issue, herein, we report an aptamer‐functionalized fractal gold nanoflower (Apt‐AuNF)‐based sensing platform for metabolite analysis. A series of fractal AuNFs with tunable nanoshell structures was synthesized using DNA‐engineered strategy, and optimized Apt‐AuNFs were demonstrated to enable efficient trap of target analytes with ∼5‐fold enhanced enrichment, thus achieving selective analysis of metabolites in complex biofluids without requirement of pre‐enrichment and purification. We applied Apt‐AuNF‐assisted laser desorption/ionization mass spectrometry for disease diagnosis, identifying diabetic patients through daily monitoring and glucose quantification in serum. This work provides a guideline to design materials for high‐performance metabolic analysis and precise disease diagnosis.