Pregnancy influences the selection of appropriate reference genes in mouse tissue: Determination of appropriate reference genes for quantitative reverse transcription PCR studies in tissues from the female mouse reproductive axis.

Selecting stably expressed reference genes which are not affected by physiological or pathophysiological conditions is crucial for reliable quantification in gene expression studies. This study examined the expression stability of a panel of twelve reference genes in tissues from the female mouse reproductive axis and the uterus. Gene expression studies were carried out using reverse transcriptase quantitative polymerase chain reaction (RT-qPCR). cDNA was synthesised from RNA extracted from hypothalami, pituitaries, ovaries and uteri of female mice at ages representing weaning, puberty and adulthood as well as pregnancy (13 ± 1 days post-coitus) (n = a minimum of 3 at each age and at pregnancy). The reference genes examined included 18 s, Actb, Atp5b, B2m, Canx, Cyc1, Eif4a2, Gapdh, Rpl13a, Sdha, Ubc and Ywhaz. The RT-qPCR raw data were imported into the qBASE+ software to analyse the expression stability using GeNorm. These data were also subsequently analysed using other software packages (Delta CT, Normfinder, BestKeeper). A comprehensive ranking was conducted considering all stability rankings generated from the different software analyses. B2m and Eif4a2 deviated from the acceptable range for amplification efficiency and therefore were excluded from the further analyses. The stability of the reference genes is influenced by the software used for the analysis with BestKeeper providing markedly different results than the other analyses. GeNorm analysis of tissues taken at different ages but not including pregnant animals, indicated that the expression of the reference genes is tissue specific with the most stable genes being: in the hypothalamus, Canx and Actb; in the pituitary, Sdha and Cyc1; in the ovary, 18s, Sdha and Ubc; and in the uterus, Ywhaz, Cyc1, Atp5b, 18s and Rpl13a. The optimal number of reference genes to be used was determined to be 2 in the first three tissues while in the uterus, the V-score generated by the GeNorm analysis was higher than 0.15 suggesting that 3 or more genes should be used for normalisation. Inclusion of tissues from pregnant mice changed the reference genes identified as being the most stable: Ubc and Sdha were the most stable genes in the hypothalamus, pituitary and the ovary. The addition of pregnant tissue had no effect on the stability of the genes in uterus (Ywhaz, Cyc1, Atp5b, 18s and Rpl13a). Identification of these stable reference genes will be of use to those interested in studying female fertility and researchers should be alert to the effects of pregnancy on reference gene stability. This study also signifies the importance of re-examining reference gene stability if the experimental conditions are changed, as shown with the introduction of pregnancy as a new factor in this research.