Antibody-Free SPR Detection of Human Myoglobin in Serum by a Sandwich Configuration of Epitope-Imprinted Nanofilms and Nanoparticles.

Abstract

Designing robust, fully synthetic receptors capable of selective protein detection remains critical for advanced clinical diagnostics. Molecularly imprinted polymers (MIPs) are a promising alternative to antibodies, but their application is often limited by empirical epitope selection and incomplete integration into functional assays. Here, we present an end-to-end platform combining data-driven epitope selection with polynorepinephrine-based molecular imprinting to produce fully synthetic protein receptors. The Python script "Epitope Selection Rational Approach" (ESRA), integrating physicochemical descriptor analysis and supervised machine learning, was developed and applied to human myoglobin (MYG) to identify effective imprintable peptide epitopes in the pre-analytical stage. Five selected epitopes were used to fabricate molecularly imprinted PNE nanofilms (MIPNE-NFs) and nanoparticles (MIPNE-NPs), targeting distinct regions of MYG. Systematic kinetic and affinity characterization by surface plasmon resonance (SPR) revealed pronounced epitope- and format-dependent recognition, underscoring the need for parallel evaluation of nanofilm and nanoparticle architectures. The optimal NF/NP combination, exploited as a capturing probe and a signal enhancer, respectively, was implemented in a fully antibody-free SPR sandwich assay. This configuration achieved sensitive detection over 4-125 ng mL-1, with a limit of detection of 0.9 ± 0.5 ng mL-1 in buffer and 0.70 ± 0.02 ng mL-1 in 1/200 diluted serum, covering the clinically relevant range for acute myocardial infarction, rhabdomyolysis, and muscle injury. This work establishes a generalizable workflow, from rational epitope selection to imprinting, kinetic validation, and assay integration, demonstrating that MIPNE can provide robust, antibody-free alternatives for sensitive protein detection in complex biological matrices.