The increasing incidence of prostate cancer (PCa) necessitates precise diagnosis and treatment to improve patient outcomes. Current diagnostic tools, such as PSA tests, MRI and biopsy, have limitations in specificity, cost, and invasiveness. Also, the available PCa targeted therapies and their effectiveness are limited. One of the common reasons for both is oversimplified response to biological signals and difficulty in dealing with the complex tumor environment. Nanotechnology holds promise for precision medicine, but intelligent systems capable of autonomous analysis and response to tumor complexity are still needed for refined PCa diagnosis and treatment.

First, machine learning identified PCa-specific targets, forming a 3-dimensional (3D) diagnostic panel at the DNA-RNA-protein level. A molecular encoding machine was then constructed to convert the various biological signals into a unified output. This output was synthesized and reported by a nucleic acid logic device, which simulates the computation of traditional silicon-based chips. Finally, the CRISPR-Cas system/nano-framework was employed to create a diagnostic and delivery platform activated by upstream signals to enable precision diagnosis and therapy for PCa. Blood and tissue samples from 36 pairs of PCa and BPH patients were collected for diagnostic tests.

A 3D diagnostic panel was developed to include GSTP1 methylation, miR-153, miR-183, and PSA. A Three-way Junction-incorporated Double Hairpin molecular encoder was engineered to unify signals, facilitating the encoding and autonomous computation of up to eight molecular signals, with a signal-to-noise ratio of 320. The CRISPR-Cas sensitively amplified signals with a detection limit down to 0.01%. The intelligent nanomachine achieved a diagnostic sensitivity of 98% and specificity of 97% in tissue samples, and sensitivity of 90% and specificity of 87% in blood samples. Additionally, a DOX-loaded nucleic acid framework-based delivery system was established, selectively targeting and eradicating tumors.

We present a new diagnostic and therapeutic platform for PCa using intelligent nucleic acid nanomachines. This platform enables detailed tumor analysis and targeted responses, offering high sensitivity, biocompatibility, and cost-effectiveness. The findings may support new strategies in precision PCa diagnosis and treatment.