An investigation into the biology and medicinal properties of Eucomis species.

Thesis (Ph.D.)-University of Natal, Pietermaritzburg, 1999.

Eucomis (Family Hyacinthaceae) are deciduous geophytes with long, narrow leaves and erect, densely packed flower spikes. The bulbs are greatly valued in traditional medicine for the treatment a variety of ailments, and are thus heavily harvested for trade in South Africa's "muthi" markets. Eucomis species propagate relatively slowly from offsets and seed, and this, together with their over-utilization ethnopharmacologically, has led to their threatened status. This investigation focussed mainly on the study of the anti-inflammatory activity of plant extracts prepared from the leaves, bulbs and roots, and the development of suitable tissue culture protocols for the bulk propagation of the species under study. Common underlying symptoms in the majority of ailments treated with traditional remedies prepared from Eucomis species are pain and inflammation. Prostaglandins are the primary mediators of the body's response to pain and inflammation, and are formed from essential fatty acids found in cell membranes. This reaction is catalysed by cyclooxygenase, a membrane-associated enzyme occurring in two isoforms, COX-1 and COX-2. Non-steroidal anti-inflammatory drugs (NSAIDs) act by inhibiting the activity of COX. The use of commercially available COX-1 inhibitors is associated with side-effects, including gastric and renal damage. Selective COX-2 inhibitors do not have these undesired effects, and are thus potentially very valuable to the pharmaceutical industry. The relative inhibitory effects of different extracts of Eucomis species on the activities of purified cyclooxygenase enzyme preparations (COX-1 in sheep seminal vesicles, COX-2 in sheep placenta) were assessed. The COX-1 assay was used to screen extracts from 10 species of Eucomis and one hybrid species at a concentration of 250 μg mℓ ¯¹ in the assay. High levels of anti-inflammatory activity were exhibited by the ethanolic extracts prepared from the dried leaves, bulbs and roots. Aqueous extracts (screened at 500 μgmℓ ¯¹) showed much lower levels of activity. In general, the highest levels of anti-inflammatory activity were observed for the ethanol bulb and root extracts. Comparison of the activity of the bulb extracts from bulbs harvested in summer and winter revealed very little difference in COX-1 inhibitory activity. Eucomis extracts were separated using thin layer chromatography. The plates were developed in a solvent system of benzene : 1,4-dioxan : acetic acid, 90:25:4 and stained with anisaldehyde-sulphuric acid. The TLC fingerprints prepared from these extracts showed different chemical profiles for the leaf, bulb and root extracts, but many similarities between the different species. The position of the active R[f] fractions was determined and correlated with the TLC-fingerprints. The most widely utilized species medicinally, E. autumnalis subspecies autumnalis, was chosen for further investigation. The fluctuation of anti-inflammatory activity with season and physiological age was determined. Young plants were found to have high levels of COX-1 inhibitory activity, particularly in the leaves. As the plant matured, higher levels of activity were associated with the bulb and root extracts. The antiinflammatory activity of the leaf, bulb and root extracts varied slightly throughout the year, with the highest levels detected towards the end of the growing season, shortly before the onset of dormancy. This study of E. autumnalis autumnalis was extended to investigate the effects of environmental conditions on the levels of COX-1 inhibitory activity. The extent to which high temperature and light intensity, fertilization of the plants in summer with Kelpak preparations, and cold storage of the dry bulbs during winter, affected the levels of active compounds accumulated, was determined. Kelpak application decreased the anti-inflammatory activity of the leaf, bulb and root extracts, while high temperature / high light intensity had no significant effect on the COX-1 inhibitory activity of the leaf or bulb extracts. The root extract did show a significant increase in anti-inflammatory activity. Bulbs that were removed from the soil and stored at 10°C exhibited significantly higher COX-1 inhibitory activity than the control bulbs maintained in the soil. Higher COX-1 inhibition was observed in the leaf extracts from these plants when harvested half-way through the growing season. No significant difference was observed at this stage between the bulb and root extracts from the different treatments. Bioassay-guided fractionation (using the COX-1 assay) was used to isolate the active principle(s) in the bulb extract. The bulb material was subjected to serial extraction using a Soxhlet apparatus. The ethyl acetate fraction showed the highest levels of COX-1 inhibition, and this was further fractionated using a Sephadex LH-20 column and a solvent system of cyclohexane : dichloromethane : methanol (7:4:1). The most active fraction from this separation was then purified using semi-preparative TLC and HPLC. The primary compound eluting in this fraction had an IC₅₀ value of 14.4 μgmℓ ¯¹ in the COX-1 assay, and 30.5 μgmℓ ¯¹ in the COX-2 assay. This compound was tentatively characterized as a phenol ring attached to a conjugated hydrocarbon chain (with a molecular weight of 390), and was a potent COX-1 inhibitor. The COX-2 / COX-1 inhibitory ratio was calculated to be 2.1. A second, highly active compound, with IC₅₀ values of 25.7 μgmℓ ¯¹ and 21.8 μgmℓ ¯¹ in the COX-1 and COX-2 assays respectively, crystalized from one of the Sephadex LH-20 column fractions. This compound was identified as a spirostane-type triterpenoid, eucosterol, previously isolated from Eucomis species but not specifically linked to the pharmacological activity of the extracts. This compound showed COX-2 / COX-1 inhibitory ratio of 0.8, indicating that it was a selective COX-2 inhibitor. Two further compounds were identified from this extract, after crystallization from different fractions obtained from Sephadex LH-20 chromatography. These were both homoisoflavanones, 5,7-dihydroxy-6-methoxy-3-(4-methoxy benzyl)-chroman-4-one, and 5,7-dihydroxy-3-(4-methoxy benzyl)-chroman-4-one [eucomin], the latter having been isolated previously. The first compound exhibited very low levels of both COX-1 and COX-2 inhibition, and the second compound (eucomin) exhibited high COX-1, but low COX-2 inhibitory activity. The in vitro propagation of the genus Eucomis was undertaken primarily to provide a source of material for experimentation, and also to optimize this technique for the bulk production of plants for commercial and conservation purposes. Multiple shoot production was initiated from leaf explants, in all species studied. A Murashige and Skoog (MS) medium, supplemented with 100 mg ℓ ¯¹ myo-inositol, 20 g ℓ ¯¹ sucrose, and solidified with 2 g ℓ ¯¹ Gelrite® was used. The optimal hormone combination for shoot initiation in the majority of species was determined to be 1 mg ℓ ¯¹ NAA and 1 mg ℓ ¯¹ BA. Optimal root initiation was demonstrated on media supplemented with 1 mg ℓ ¯¹ IAA, IBA or NAA, depending on species. A continuous culture system using this protocol produced 25-30 plantlets per culture bottle, with 10-25 specimens per bottle available for acclimatization. To maximize plantlet survival, different support media used during the acclimatization process were necessary. Certain species responded best on a vermiculite medium, while perlite (which holds less water) was necessary for the optimal survival rate of other species. Acclimatized plantlets were repotted in a sand : soil mix (1:1). Further experimental work aimed to determine the factors affecting the accumulation of anti-inflammatory compounds in in vitro plantlets. Extracts prepared from in vitro plantlets showed high levels of COX-1 and COX-2 inhibitory activity, with a C0X-2/C0X-1 ratio of 1.1. High levels of sucrose (40 g ℓ ¯¹) significantly increased the number of shoots initiated, but had no effect on the anti-inflammatory activity. Low levels of sucrose (10 g ℓ ¯¹) led to a significant decrease in COX-1 inhibition. Changing the levels of nitrogen in the medium (but not the ratio of nitrate to ammonium ions) had no significant effect on the COX-1 inhibitory activity of the extracts. Callus was initiated from leaf explants and experiments were conducted to maximize callus proliferation. Optimal callus growth occurred on an MS medium supplemented with 100 mg ℓ ¯¹ myo-inositol, 30 g ℓ ¯¹ sucrose, 2 g ℓ ¯¹ Gelrite® , and a hormone combination of 10 mg ℓ ¯¹ 2,4-D and 2 mg ℓ ¯¹ kinetin. Callus cultures maintained in the dark grew best. Callus extracts tested in the COX assays (250 μgmℓ ¯¹) showed a higher level of COX-2 inhibition (69%) than COX-1 inhibition (46%). Lastly, the conclusive identification of the species under study was attempted, using DNA fingerprinting. Protocols were developed for the extraction of DNA from the leaves of Eucomis plants, and the optimization of the AP-PCR technique. Random sequence (10-base) oligonucleotide primers were screened, each primer used singly. Primers were selected on the basis that more than five distinct bands were detected. Differences were detected in the amplification products visualized using nondenaturing agarose gel electrophoresis stained with ethidium bromide. This work provides the basis for further studies into the phylogenetic relationships between the various species (and hybrids) of Eucomis.