Screening foods of the endangered Kakapo parrot ( Strigops habroptilus ) for oestrogenic activity using a recombinant yeast bioassay

Abstract

In the past decade the possible effects of environmental oestrogens, both natural and synthetic, on the reproduction of vertebrates has become an issue of both public and scientific interest (Turner and Sharpe, 1997). While the general significance of such chemicals remains controversial there is clear evidence that, in some contexts, phytooestrogens can influence the fertility of vertebrates (Leopold et al., 1976; Shutt, 1976). While the impact of natural and synthetic environmental oestrogens may be slight, highly endangered species do represent a unique situation in which even modest reductions in the fertility of key individuals may have implications for the survival of the entire species. From these considerations it was deemed prudent to determine if oestrogenic activity could be detected in the natural and artificial foods of the endangered New Zealand nocturnal parrot, the Kakapo (Strigops habroptilus; Powlesland and Lloyd, 1994). Two issues were to be addressed: (i) whether physiologically significant levels of oestrogenic activity were detectable in the artificial (adult supplementary and chick-raising) foods of Kakapo; and (ii ) whether oestrogenic activity could be detected in either crude extracts of the Kakapo’s natural foods (i.e. native tree fruits) or chemicals isolated from these foods. A simple and robust recombinant yeast bioassay (developed by Glaxo-Wellcome) was adapted for this work. Briefly, the recombinant yeast strain used has been transformed with two genes: the human oestrogen receptor (hER) and a lacZ reporter gene located adjacent to an oestrogen response element (ERE-lacZ). Transcription of the lacZ gene in this strain is regulated through the ERE; thus the cells produce increased amounts of b -galactosidase enzyme (encoded by lacZ ) in response to any chemical that can activate the hER and cause it to bind the ERE (5). It should be emphasised that application of this bioassay requires little prior knowledge of the chemistry properties of possible oestrogenic chemicals other than requiring their solubility in 100–70% ethanol. A typical extraction procedure for the foods was as follows: weighed samples (c.100 mg) were extracted overnight (30°C on a rotating wheel) in 1·5 ml of either water alone or water with b -glucosidase (10 or 100 μg/ml). Isoflavones are present in plant cells largely as conjugated glucosides which show little oestrogenic activity: therefore b -glucosidase was added to some extraction suspensions to catalyse hydrolysis of sugar– isoflavone bonds. After adding 3·5 ml of absolute ethanol (i.e. making the extraction solvent to 70% ethanol) the samples were again extracted overnight, centrifuged and filtered through a 0·22 μm filter. Dilution series of 10 μl aliquots of the extracts were assayed as described in Routledge and Sumpter (1996). Foliage of oestrogenic clovers was used as a positive control for both the extraction and assay procedure. Purified phyto-chemicals assayed were dissolved in 96% ethanol before a dilution 1:1 series was assayed. No oestrogenic activity was detected in 70% ethanol extracts of the following adult supplementary foods: sunflower seeds, safflower seeds, walnuts, pumpkin seeds and almonds. However, very low levels of oestrogenic activity were detected in extracts of one of three commercial chick-raising products screened. The bioassay detected high levels of oestrogenic activity in 70% ethanol extracts of the positive control, clover leaves. The following chemicals, possibly present in fruits forming part of the natural Kakapo diet, were screened using the bioassay: coniferyl alcohol, gibberellins (GA9, methyl-GA9, GA4, methyl-GA4, GA3, methyl-GA3), totarol, kaurene, kaurenoic acid, phyllocladene, brassinolide. No oestrogenic activity was detected for any of these chemicals. An earlier paper (Brandt and Ross, 1948) had reported (without presenting data) oestrogenic activity for the alcohol derivative of podocarpic acid, a major component of the resin of rimu (Dacrydium cupressinum), a New Zealand forest tree that is a major natural food source for Kakapo. We confirmed this report, showing podocarpinol had oestrogenic activity comparable with genistein (c.10-fold less than 17-b estradiol) while podocarpic acid itself also had weak oestrogenic activity (c.10-fold less than 17b estradiol). The significance, if any, of this to the reproduction of Kakapo is not clear.