SERS ratiometric strategy leveraging carbon allocation trade-off enables Tremella fuciformis metabolic phenotyping and traceability

Accurate metabolic phenotyping and geographical traceability of gelatinous fungi are critical for uncovering fungal physiological adaptation mechanisms and advancing edible fungus quality supervision. However, high phenotypic similarity and subtle metabolic differences in homologous fungal samples pose formidable challenges to conventional physicochemical assays. Especially, relying on chemical fingerprinting without biological phenotypic implications, conventional surface-enhanced Raman scattering (SERS) analysis is unable to discern delicate metabolic discrepancies in fungal samples.

Herein, we propose a metabolic characteristic peak ratio-based SERS strategy using Tremella fuciformis (T. fuciformis) as the model. By extracting intrinsic metabolic signatures represented by the carbon allocation trade-off between stress-resistant functional polysaccharides and storage-type carbohydrates, it delivers metabolic specificity and detection stability/reproducibility that are not possible with traditional fingerprint-based SERS fungal analysis. Integrated with machine learning, this method accurately differentiates six T. fuciformis samples, including specimens with distinct metabolic phenotypes as well as geographically differentiated samples with highly similar phenotypic features, achieving a superior accuracy of 99.2% and an AUC of 1.0. Combined with a custom-built portable Raman device we constructed, the entire detection procedure can be completed within 15 min, enabling efficient on-site analysis.

This study establishes a bio-signature-oriented SERS paradigm that upgrades chemical fingerprint identification to fungal metabolic phenotyping, providing a versatile technical platform for exploring fungal metabolic variation and phenotypic differentiation.