Session 8

Chayanis Sutcharitchan¹, Worakorn Wiwatcharakornkul¹, Somchai Keawwangchai², Tasawan Kaewwangchai², Sornkanok Vimolmangkang¹, Suchada Sukrong¹, Wanchai De-Eknamkul<sup>1

1</sup>Department of Pharmacognosy and Pharmaceutical Botany, Faculty of Pharmaceutical Sciences, Chulalongkorn University, Bangkok 10330, Thailand
²Department of Chemistry, Faculty of Science, Mahasarakham University, Mahasarakham 44150, Thailand

This email address is being protected from spambots. You need JavaScript enabled to view it.

Thailand is known to be one of the world’s best sources of cannabis strains, and thus has a potential to be a cannabis hub in Asia. Thai strains of cannabis commonly refer to 100% sativa strains native to Southeast Asia. Each strain contains different ratios of the psychoactive substances, tetrahydrocannabinol (THC) and cannabidiol (CBD), which are known to have medical benefits. Three major cannabis chemotypes (THC dominant, intermediate, and CBD dominant) and two minor ones (CBG dominant and cannabinoid-free) have been categorized for chemotaxonomic discrimination. So far, very little has been known about the chemotypes and genetic backgrounds of the diversified Thai cannabis strains and hybrids. Therefore, the techniques of HPTLC and DNA profilings were introduced to observe the relationship between their metabolic and genetic levels. Almost 100 samples of dried flowers obtained from various strains of Thai cannabis and their hybrids were collected and prepared for HPTLC and DNA analyses. The solvent system used for the HPTLC separation of cannabis extracts was modified from the previous report (1). The results showed that there was a high variation in the chemotypes among the Thai cannabis samples. For the DNA analysis, tetrahydrocannabinolic acid-synthase (THCA-synthase) and cannabidiolic acid-synthase (CBDA-synthase) genes (2,3) were used as the chemotype markers of various cannabis samples. The preliminary results obtained by PCR and DNA sequencing techniques suggested that there was a good correlation between the HPTLC chemical profiles and the presence of functional forms of THCA-synthase (drug-type) and CBDA-synthase (fiber-type). Therefore, the sequence of both genes can be used for early prediction of the cannabinoid profiled in Thai cannabis plants.

Acknowledgements
This work was supported by Chulalongkorn University's Natural Product Biotechnology Research Unit (for WD) and Center of Excellence in DNA Barcoding of Thai Medicinal Plants (for SS).

References
[1]  CAMAG Application Note A-108.1 (2017).
[2]  C. Onofri, E. de Meijer, G. Mandolino, Phytochemistry 116 (2015) 57–68.
[3]  A. R. Garfinkel, M. Otten, S. Crawford, Genes 12 (2021) 228. https://doi.org/10.3390/genes12020228

Bernd Spangenberg

University of Applied Sciences, Offenburg, Badstrasse 24, 77652 Offenburg, Germany

This email address is being protected from spambots. You need JavaScript enabled to view it.

A light-emitting diode (LED) is a semiconductor light source emitting light when current flows through it. Since the Nobel Prize-winning invention of high brightness blue LEDs in 1993 by Shuji Nakamura, LEDs with excitation wavelengths below 400 nm have been suitable illumination sources for fluorescence measurements. In high-performance thin-layer chromatography (HPTLC), these devices open up a new field in planar chromatography because their long-term stable light intensity makes them a valuable light source for fluorescence analysis directly on the plate. Examples of fluorescence measurements in HPTLC using LEDs of different wavelengths in combination with diode-array planar chromatography will be presented. Some reasons are given why fluorescence measurements should be preferably performed on RP-18 plates and not on uncoated silica gel plates.

References:
Spangenberg, C.F. Poole, C. Weins. Quantitative Thin-Layer Chromatography: A Practical Survey, Springer, 2011

Matthias Guggenberger¹, Josua T. Oberlerchner¹, Heinrich Grausgruber², Thomas Rosenau¹, Stefan Böhmdorfer<sup>1

1</sup>University of Natural Resources and Life Sciences, Vienna (BOKU), Institute of Chemistry of Renewable Resources, Konrad-Lorenz-Straße 24, 3430 Tulln, Austria
²University of Natural Resources and Life Sciences, Vienna (BOKU), Institute of Plant Breeding, Konrad-Lorenz-Straße 24, 3430 Tulln, Austria

This email address is being protected from spambots. You need JavaScript enabled to view it.

A reliable, chemometric evaluation of HPTLC chromatograms for identification and classification would be most useful to screen larger datasets for unusual samples, or to verify the identity of unknown samples with low operator load. Typically, linear models are used for these tasks, with limited success due to the high level of noise in visual detection and inadequate plate-to-plate reproducibility.
A Self Organizing Map (SOM) is an iterative, unsupervised classification algorithm. We applied it to two sets of HPTLC data (495 analyses of essential oils, 50 analyses of wheat anthocyanins; Vis detection). The sample sets were sorted correctly (97% of the essential oils); grouping by plate was never observed. Misplaced chromatograms were either distorted by a volatile component of the neighboring track or contained an admixture. The impaired samples provided insight as to method development and validation. In the anthocyanin set, misplaced samples were always indicating wheat samples with unusual provenance.
To classify unknown samples, two extensions were successfully used: using a global SOM of all training samples, or training one SOM for each class. Both cases allow rating the confidence of the classification by a quality factor. This can also be used to test robustness, as deficient plates will show a lower match quality.
Data selection (color channels, color spaces) and preprocessing was investigated. The method proved sensitive to variable migration distances, thus chromatography had to be conducted meticulously.
Thus, self-organizing maps permit an automated classification of HPTLC chromatograms with a safeguard, as match quality is indicated. They also allow to efficiently select unusual samples from a larger data set for a more detailed (and costly) analysis.

Acknowledgments
We are grateful for the essential oils provided as a gift by Sonnentor Kräuterhandelsgesellschaft mbH and for the support of the Austrian Biorefinery Center Tulln (ABCT).

Reference
[1] M. Guggenberger, J. T. Oberlerchner, H. Grausgruber, T. Rosenau, S. Böhmdorfer, Talanta 233 (2021), 122460.