A high-throughput multivariate statistics platform for the discovery of tyrosinase inhibitors

The implementation of high-throughput screening (HTS) technologies has become an indispensable tool for the detection of bioactive constituents, avoiding re-isolation of known compounds, reducing workload and cost. The aim of our study was the establishment of an integrated high-throughput multivariate statistics platform, relying on FCPC, HPTLC, and NMR, for the direct detection and identification of natural compounds with skin whitening properties prior to any isolation. Greek flora - due to its high biodiversity - was used as a source for the collection of selected plant material. Previous in vitro investigation of plant extracts against their tyrosinase inhibition activity revealed nine extracts as the most promising. The selected extracts were fractionated by FCPC using a certain protocol and each fraction was split in 3 equal parts for: i) HPTLC profiling and bioautography, ii) NMR profiling iii) in vitro assay. An integrated HPTLC-based procedure for the tracing of compounds that contributed to tyrosinase inhibitory effect in active fractions was established with the use of multivariate data analysis [1]. Additionally, NMR spectral data were correlated with the results of tyrosinase activity resulting in the identification of bioactive compounds through the combination of the Heterocovariance approach (HetCA) [2] and the statistical total correlation spectroscopy (STOCSY) [3]. The combined data deriving from NMR and HPTLC correlated to the biological activity by the statistically driven approach, revealed flavans, flavonols, phenolic compounds and stilbenoids between the most promising whitening agents, providing a major reduction in workload by direct use of routine information.

• References

• 1 Chaita E, Gikas V, Aligiannis N. Integrated HPTLC-based methodology for the tracing of bioactive compounds in herbal extracts employing multivariate chemometrics. A case study on Morus alba . Phytochem Anal 2017; 28 (02) : 125-131
• 2 Aligiannis N, Halabalaki M, Chaita E, Kouloura E, Argyropoulou A, Benaki D. et al. Heterocovariance based metabolomics as a powerful tool accelerating bioactive natural product identification. ChemistrySelect 2016; 1: 2531-2535
• 3 Boka VI, Stathopoulou K, Benaki D, Gikas E, Aligiannis N, Mikros E, Skaltsounis LA. Could multivariate statistics exploit HPTLC and NMR data to reveal bioactive compounds? The case of Paeonia mascula . Phytochem Lett 2017; 20: 379-385