D. Michael Fry*, MS Denison, B Puschner, ML Johnson, I Werner, and SJ Teh.

*Department of Animal Science, University of California

There are two basic approaches to identification and monitoring of endocrine disrupting substances in the environment. The USGS has initiated a monitoring program of US rivers and other surface waters for "emerging pollutants", which include antibiotics, hormones, and pharmaceuticals. 95 chemicals are being monitored, including 17 chemicals and hormones that have the potential of disrupting reproductive hormones or development. The monitoring program has the ability to identify levels of these endocrine disruptors in watersheds, but the program does not directly address the activity of any other compounds, or the potential for endocrine disruptive activity as side effects of other chemicals in the environment.

We have developed an alternative strategy to collect and screen water samples for hormonal activity (estrogenic, anti-estrogenic, androgenic, and anti-androgenic), and identify the active chemicals in water samples. The program is also designed to identify unknown chemicals with endocrine disruptive activity, and will quantify the activity levels in surface water samples, to ultimately provide guidance on risk of surface waters to humans, fish and wildlife.

The program is a four step set of analyses: 1) Water sampling and determination of hormonal activity relative to estradiol or testosterone in reference binding assays; 2) Identification and quantification of the presumptive active chemical(s) in the sample;
3) Hormonal activity and developmental testing of identified chemicals in a fish test species (Gambusia affinis) common throughout California watersheds; and 4) Comparison of the hormonal activity of the identified chemical with watershed environmental levels, to determine the risk posed to drinking water and fish.

About 1000 water samples from central California waters will be tested each year. Water sampling will be conducted using a protocol similar to that of Snyder et al. (2001a,b), in which 5 liter water samples are pumped through solid phase extraction cartridges or disks in the field. Each sample cartridge/disk will be identified in the field with a bar-code label, with data entered into a GPS database in a laptop computer in the field.

Extracts of each sample will be tested for hormonal activity using a recombinant human ovarian carcinoma (BG-1) cell line (called BG1Luc4E2 cells) that contains a stably transfected estrogen-responsive firefly luciferase reporter gene. These cells respond in a time, dose and chemical specific manner-dependent manner to 17ßestradiol and related xenoestrogens (including o'p'-DDT, methoxychlor, kepone, bisphenol a, and nonylphenol) with up to a 100-fold induction of luciferase activity. In addition, this cell bioassay is extremely sensitive, with a lower limit of detection of at least 0.1 pM of estradiol (Rogers and Denison 2000). In addition to estrogenic activity, we will also examine the androgenic activity of the same samples using a recombinant prostate cell line (LNCaP) that has been transiently transfected with an androgen-responsive luciferase reporter gene.

Up to 50 of the highest activity water sample extracts will be analyzed each year by one or several analytical techniques and the chemical(s) present will be identified by mass spectroscopy, and compared to analytical standards. The hormonal activity of the identified chemicals will be confirmed in the cell culture assays using analytical grade compounds.

We will select the most significant compounds for fish analysis. We plan to conduct toxicological assessments with 12-15 compounds during the first 3 years. We anticipate that a larger number of hormone active chemicals will be identified in the watershed than will be possible to be tested on Gambusia, because of the time and labor-intensive nature of the fish screening and biomarker assays. Gambusia will be exposed to analytical grade compounds using the techniques adapted from the study of Koger, Teh, and Hinton (2000) examining gonadal morphology, fertility, and embryo/larval viability. The numbers of live young, sex ratios of surviving progeny, and histological evaluation of gonads will be used as markers of reproductive function. Detection of the specific protein biomarkers vitellogenin and choriogenin will be used as biomarkers for detecting hormonal activity in adult Gambusia.

The dose response information gathered from the fish exposure studies will be compared to the environmental concentrations of active chemicals in the watershed, to determine the environmental health risk for fish and drinking water. For agricultural chemicals, the water sample locations will be compared with the Department of Pesticide Regulation use reports to identify other likely hot spots within the watershed.