Transcript variability and physiological correlates in the fathead minnow ovary: Implications for sample size, and experimental power.

Fundamental studies characterizing transcript variability in teleost tissues are needed if molecular endpoints are to be useful for regulatory ecotoxicology. The objectives of this study were to (1) measure transcript variability of steroidogenic enzymes and steroid receptors in the fathead minnow (FHM; Pimephales promelas) ovary to better determine normal variability and the sample sizes needed to detect specific effect sizes and to (2) determine how expression patterns related to higher level endpoints used in some regulatory ecotoxicology programs (e.g. relative gonad size). Estrogen receptor 2b (esr2b) and 5α-reductase a3 (srd5a3) showed high variability in the ovary (CV>1.0) while progesterone receptor (pgr), androgen receptor (ar), and esr2a showed comparatively low variability (CV=~0.5--0.7). Using these estimates, a power analysis revealed that sample sizes for real-time PCR experiments would need to be>20 to detect a 2-fold change for 7 of the transcripts examined; thus many molecular studies conducted in the fish ovary may have insufficient power to detect smaller effects. Two transcripts were correlated to steroid production in the ovary; cyp19a1 levels were positively correlated to in vitro E2 production, while ar levels were negatively correlated to in vitro T production. Thus, these transcripts may be informative molecular surrogates for ovarian steroid production. No transcript investigated showed any correlation to GSI, condition, or body weight/length. Molecular approaches in fish are increasingly used to assess biological impacts of chemical stressors; however additional studies are required that determine how molecular variability relates to higher level biological endpoints.