SERS-based spectral analysis of ITS1 PCR products for differentiation of three old world Leishmania species.

Leishmaniasis, a zoonotic disease caused by parasites of the genus Leishmania, poses a significant medical and veterinary importance worldwide. This study was designed to explore the potential of SERS-based plasmonic substrate combined with advanced multivariate statistical analysis for differentiation of three prevalent Old World Leishmania species. In the present study, we investigated the potential of Raman spectroscopy and surface-enhanced Raman spectroscopy (SERS) combined with machine-learning-based analytical approaches for differentiation of three prevalent Old World Leishmania species. Previously characterized ITS1 PCR products corresponding to Leishmania infantum, Leishmania tropica, and Leishmania major were used as target molecules for spectroscopic analysis. Species identity of the samples had been confirmed previously using ITS1 PCR, sequencing, and PCR-RFLP analysis. Raman and SERS spectra were acquired using a HORIBA Raman microspectrometer equipped with a He-Ne laser. Silver nanoparticle-based SERS substrates were synthesized using a modified Tollens' method to enhance Raman signal intensity. Spectral datasets were preprocessed and analyzed using principal component analysis (PCA), combined with linear discriminant analysis (PCA-LDA). The PCA-LDA technique reduced data dimensionality and facilitated visualization of spectral separation between species. A support vector machine (SVM) classifier was also utilized as an exploratory approach to visualize potential decision boundaries among species based on the SERS data. SVM classifiers on the ITS1 product spectra suggested the potential ability of SERS-derived spectral features to distinguish among three prevalent Old World Leishmania species. The obtained classifications were fully consistent with the previously confirmed molecular identification results. The findings demonstrate that SERS coupled with multivariate statistical analysis and machine-learning approaches can provide a promising analytical framework for species-level differentiation of Leishmania parasites based on ITS1 PCR products. This preliminary study highlights the potential application of Raman/SERS-based molecular profiling for future diagnostic and epidemiological investigations. Broader sampling across geographically diverse isolates and independent validation datasets will be required in future investigations to determine the reproducibility, specificity, and potential translational value of this analytical approach.