Steven D. Christenson and Drs. Gary M. Booth and Steven G. Wood
Vai Haka is the Tongan name given to a medicinal extract prepared and prescribed by indigenous healers on the island kingdom of Tonga. Used as a treatment for morning sickness (hyper-emesis), the mixture is prepared by boiling the barks of four different rainforest tree species. These species are Tarrena sambucina, Syzigium malaccense, Alphitonia zizyphoides, and Macaranga harveyana. Studies by Cox (et al) demonstrated biological activity in each of these species from a Guinea Pig Ileum assay.1 Further work by Cox documented that each species is used individually by Samoan healers to treat internal distress.2 Thus bioactivity is well documented through both modern science and indigenous tradition. Given its demonstrated bioactivity it is useful to further screen for bioactivity with the intent of identifying the active agents as a source of new drugs.
Although the medicine was first distinguished by its antiemetic properties these proved impossible to duplicate in the laboratory due to lack of a proper test organism. To overcome this difficulty the medicine was screened on a variety of different assays to identify a practical indicator of bioactivity. Several different toxicity assays on fungus and bacteria were tried and showed no significant activity. A final attempt using cancerous uterine (HeLa) tissue cultures did show significant activity. This activity was characterized by using an MTT assay. The MTT assay measures changes in cellular metabolism by replacing oxygen with another oxidative agent that possesses photometric activity. Not only is such activity interesting due to its implications for cancer treatment, but also ideal for qualitative and quantitative comparisons.
With HeLa cell fine activity as a guide six different fractions were tested for their biological activity. These fractions were prepared through the use of both column chromatography and two phase liquid-liquid extraction. Column chromatography of the crude medicine through Sephadex LH-20 resin yielded three fractions characterized by the eluting solvent: water (H20), 50% methanol water (NleOH), and 70% acetone. Each of these fracfions was subsequently extracted with butanol as a second phase to separate the less polar components. Each of the resulting fractions was then tested on HeLa cell cultures to identify the fraction holding the compounds responsible for the observed bioactivity of the original medicine.
The most active fraction was then subjected to thin layer chromatography (TLC) and spot testing. This method produces a crude separation and allows the identification of important chemical families contained in the fraction. Well known spot tests on a variety of biactive chemical families were employed.3 High performance liquid chromatography (HPLC) was done on an analytical scale to give an estimate of relative amounts of each compound and aid in the design of a preparative fractionation scheme. Such a scheme is needed to efficiently produce sufficient quantities for use in more detailed analysis. This report only covers up to the development of a preparative method but paves the way to further work to be done.
Results from the MTT bioassay showed a 30% decrease in activity when 20 1 of crude extract was used. Subsequent trials conducted after the various fractionation schemes showed the butanol layer of the methanol column eluent to show the greatest activity. At doses on 1 mg/ml there was a 70% reduction in cellular activity. The difference between the crude medicine activity and separated fraction can be attributed to the differences in concentration. Separation of this fraction followed by spot testing yielded four blueviolet bands when treated with concentrated sulfuric acid and vanillin solutions. This is a positive identification of Saponins a group of chemicals known for their bioactivity.4 No other spray reagents yielded significant results. From this data it is concluded that the principal biological agents are Saponins. More advanced separation, assays, and structural elucidation will confirm this conclusion. Based on information from the studies above it was possible to develop an improved method for separating the MeOH/butanol fraction from the crude medicine in the higher amounts required for separation of individual compounds and structural elucidation. BPLC chromatographs proved to be very useful in comparing samples obtained from the new method with those of the analytical method.
It was found that greater quantities of the mixture could be obtained by extracting the crude with butanol first and then very little column chromatography was even needed. The new process requires fewer solvent changes and as such needs less sterilization and concentration steps which were responsible for loss of both time and material. Chromatography data for both the original and preparative schemes show essentially the same peaks and demonstrate the same activity on MTT assays.
In summary, it was found that the medicinal preparation from Tongan healers decreased activity in cervical cancer tissue significantly. This biological activity is due to a class of compounds called Saponins. A more efficient manner of extraction and preparation was developed to allow further studies to confirm Saponins as the biological agents by elucidating their structure. The data, techniques, and resources developed by this report will continue to be used in these further studies. It has also paved the way for future work on other medicinal extracts from the Pacific islands.
References
- P.A. Cox, et al. 1989. Economic Botany 43(4): 487-497 1
- P.A. Cox. 1993. Journal ofethnopharmacology 38: 181-188 2
- S. Bladt, et al. 1984. Appendix A In: Plant Drug Analysis. Springer-Verlag, Berlin 3
- K. Hostettinann and A. Marston. 1995. Saponins, Chemistry and Pharmacology of Natural Products Series, Cambridge Univ. Press, 4
Cambridge - The aid and expertise of Dr. Stephen Wood is gratefully acknowledged, without which none of this would have been possible.