As a result of bioassay-guided fractionation, betulinic acid, a pentacyclic triterpene, was identified as a melanoma-specific cytotoxic agent. In follow-up studies conducted with athymic mice carrying human melanomas, tumour growth was completely inhibited without toxicity. As judged by a variety of cellular responses, antitumour activity was mediated by the induction of apoptosis. Betulinic acid is inexpensive and available in abundant supply from common natural sources, notably the bark of white birch trees. The compound is currently undergoing preclinical development for the treatment or prevention of malignant melanoma.

Enzymatic activity mediated by recombinant human DNA ligase I (hLI), in conjunction with tannin removal procedures, has been applied to a natural-product screen involving approximately 1000 plant extracts and various pure compounds. The primary hLI activity assay involved the measurement of the amount of radiolabelled phosphate in a synthetic nucleic acid hybrid that becomes resistant to alkaline phosphatase as a result of ligation. A bioactivity-guided fractionation scheme resulted in the isolation of ursolic [IC50=100 micrograms/ml (216 microM)] and oleanolic [IC50=100 micrograms/ml (216 microM)] acids from Tricalysia niamniamensis Hiern (Rubiaceae), which demonstrated similar DNA ligase inhibition profiles to other triterpenes such as aleuritolic acid. Protolichesterinic acid [IC50=6 micrograms/ml (20 microM)], swertifrancheside [IC50 = 8 micrograms/ml(11)microM)] and fulvoplumierin [IC50=87 micrograms/ml (357 microM)] represent three additional natural-product structural classes that inhibit hLI. Fagaronine chloride [IC50=10 micrograms/ml (27 micronM] and certain flavonoids are also among the pure natural products that were found to disrupt the activity of the enzyme, consistent with their nucleic acid intercalative properties. Further analyses revealed that some of the hLI-inhibitory compounds interfered with the initial adenylation step of the ligation reaction, indicating a direct interaction with the enzyme protein. However, in all cases, this enzyme-inhibitor interaction did not disrupt the DNA relaxation activity mediated by hLI. These results indicate that, although the same enzyme active site may be involved in both enzyme adenylation and DNA relaxation, inhibitors may exert allosteric effects by inducing conformational changes that disrupt only one of these activities. Studies with inhibitors are important for the assignment of specific cellular functions to these enzymes, as well as for their development into clinically useful antitumour agents.