NMR Spectral fingerprinting of Camellia sinensis extracts: How to analyze the NMR spectra - Similarity Calculations and Differential NMR vs. Multivariate Data Analysis

NMR Spectral fingerprinting of Camellia sinensis extracts: How to analyze the NMR spectra - Similarity Calculations and Differential NMR vs. Multivariate Data Analysis.

* Denise Dagnino1, Jan Schripsema2

1 Grupo Metabolômica, Laboratório de Biotecnologia, Universidade Estadual do Norte

Fluminense, Campos dos Goytacazes, RJ, Brazil

2 Grupo Metabolômica, Laboratório de Ciências Quimicas, Universidade Estadual do Norte

Fluminense, Campos dos Goytacazes, RJ, Brazil

*dagnino@uenf.br

 

The leaves of Camellia sinensis, a small shrub or tree native to East Asia, but nowadays cultivated

in many countries, are used to prepare the popular beverage tea. Many varieties exist, e.g small leaf

Chinese or large leaf Assam and different processing methods of the plant leaves give rise to

different teas known as black, green, white, oolong and pu-erh tea. Furthermore many different

blends are used, e.g. black tea flavoured with bergamot oil, known as Earl Grey.

In the present study about seventy different types of tea were investigated. Each sample was

extracted with a two-phase extraction developed in our laboratory1, yielding a D2O and a CDCl3

extract. From each extract the 1H NMR spectrum was obtained. Generally Principle Component

Analysis (PCA) is used to determine the grouping of samples and the loading plot reveals the

datapoints that show the largest variation between the samples. Before the processing of the spectra

with PCA a binning of the spectra is performed reducing the total number of datapoints of the

spectrum to several hundreds. This procedure, although widely used, has a number of important

disadvantages. The information about minor components is lost and individual data points are

considered independent variables which in NMR is not true.

To avoid the disadvantages an alternative methodology was developed for the NMR spectral

processing using similarity calculations and differential NMR2. The two different methodologies

were compared in this study.

Different types of tea (black, white, green, and pu-erh) did show some clustering in the PCA, but

differential NMR clearly showed their differences, mainly related to different types and levels of

catechins.

With PCA black tea and Earl Grey tea did not show a clear clustering while with differential NMR

the components of the bergamot oil which is added to the Earl Grey tea are clearly visualized.

The present study clearly showed that differential NMR is an essential procedure to find low

abundant marker compounds.

Acknowledgments: CNPq, CAPES and FAPERJ for financial support.

References

1. Schripsema, J.; Dagnino, D. In: Metabolic Profiling, Ed. Theodoridis, G. et al., Ch. 13,

Springer Verlag, vol. 1738, 195-202 (2018).

2. J. Schripsema. Metabolomics, 15, 39 (2019).