The RT-qPCR data normalization performed against endogenous reference affected by experimental treatment or tissue physiological status may greatly alter interpretation of the results and lead to inaccurate and/or erroneous conclusions16. Since steroid hormones were found to modulate the transcript levels of some of the traditionally used RGs (e.g. GAPDH and 18s RNA transcription altered by estrogen and progesterone, respectively)11,18 identification of RGs unaffected by hormonal changes is required.
In our study, we evaluated the expression stability of 10 candidate RGs in placental cell cultures obtained from bovine caruncles during pregnancy. The cell cultures were subjected to a hormonal treatment of P4, E2 and PGF2α. Stability rankings generated with the use of four algorithms (geNorm, NormFinder, BestKeeper and delta Ct method) revealed that RPS9 and SUZ12 displayed the highest expression stability in the tested material. Furthermore, according to geNorm calculations, only two best performing RGs suffice to appropriately normalize RT-qPCR data. Our study also found that HPRT1 and SDHA were characterized by the highest variability out of all RGs analysed.
While studying maternal-conceptus immune tolerance in bovine endometrium Chaney et al. (2022) assayed a set of eight potential reference genes in terms of their expression stability21. For caruncular tissues treated with either progesterone (P4), recombinant bovine galectin-1 (rbLGALS1) or their combination, ring finger protein 11 gene (RNF11) and H3 histone family member 3A gene (H3F3A) were chosen as internal controls. Similarly to our results, SDHA gene was not among the most stable RGs in caruncular endometrium. Nevertheless, it proved to be among the two-best RGs stably expressed in intercaruncular bovine endometrium treated with rbLGALS1.
Likewise, SDHA along with PPIA (peptidylprolyl isomerase A gene) were identified as the most stable RGs across bovine endometrial explants subjected to various treatments (e.g. co-cultured with bovine conceptuses of different origin or exposed to interferon tau (IFNT)13. Moreover, SDHA was proposed as a good internal control in studies performed on the uterus of sows across pregnancy22. In the research on placental formation in cows carried out by Guillomot et al. (2014)23 SDHA, RPL19 and RPLP0 were chosen for the normalization of RT-qPCR data obtained from placental tissues, endometrial caruncles and the intercaruncular endometrium. By contrast, SDHA was demonstrated to display high expression variability in the uterus of both non-pregnant and pregnant mice16.
The primary function of the corpus luteum is the secretion of progesterone, which prepares the endometrium for implantation and it is essential for the maintenance of pregnancy12. The research of Rekawiecki et al. (2012) performed on corpora lutea collected from cyclic or pregnant cows revealed high expression stability of SUZ12 and CNOT11 (alternatively named C2orf29)11. This is in accordance with our results which also demonstrated high expression stability of SUZ12 and rather good expression stability of CNOT11.
In a study carried out by Mezera et al. (2020) in the bovine corpus luteum throughout early pregnancy and luteolysis a set of 74 stably expressed genes was identified, one-third of which belonged to the ribosomal protein family24. As protein synthesis is an essential process for the cell, nearly constant expression of these genes might be expected. However, the abovementioned study revealed also that RPS4X (ribosomal protein S4X) and RPL4 (ribosomal protein L4) may be used as internal controls as long as the samples were collected throughout luteolysis. In specimens coming from early pregnancy other candidates performed much better, therefore they should be chosen for data normalization. This constitutes yet another example that corroborates the need for putative RGs assessment in specific samples sets. Still, research carried out in other bovine cells and tissues confirms the potential of ribosomal protein subunit genes to serve as internal controls in RT-qPCR data analysis25–27.
Regarding the prospective use of genes encoding ribosomal protein subunits as RGs, worth mentioning is that RPL13a (ribosomal protein L13a) expression was found not to be altered by estrogen treatment in the murine uterus18. In relation to hormone-secreting organs, ribosomal genes RPS9 and RPS18 were identified as the best RGs in bovine ovaries, regardless of their luteal stages12.
Cheng et al. (2020) performed RGs selection within complex samples set of fetal tissues and maternal reproductive tissues obtained from early-pregnant cows14. The analysis of eight tissue types was carried out, including caruncular endometrium samples. According to obtained data, which was analysed by three algorithms (geNorm, NormFinder and BestKeeper), RPS9 and CNOT11 displayed the highest expression stability across all of the examined tissues. Likewise, a comprehensive ranking generated within this study points out to RPS9 as the most stable gene, whereas CNOT11 is reported to have good expression stability. In the aforementioned study, GAPDH and HPRT1 were indicated as the least stable RGs by all used algorithms. This is also in concordance with our results, which ranked HPRT1 as the worst-performing gene out of all candidates tested. However, we did not test GAPDH. On the other hand, some studies identified HPRT1 as one of the two best RGs in the porcine placenta and ovaries but not in the porcine endometria22,28,29.
As stated by Craythorn et al. (2009), the identification of appropriate RGs in uterine tissues might be a challenge due to the structural alternations occurring in response to circulating hormones17. The literature shows that both GAPDH and HPRT1 expression is modulated by estrogens in the mouse uterus18. Furthermore, their expression was reported to be altered by sex hormone exposure in human skin cells30.
Selection of RGs conducted in ovaries of riverine buffaloes (Bubalus bubalis) showed that GAPDH displayed the most variable expression out of 10 candidates analysed31. Similarly, GAPDH ranked as the least stable RG in the corpus luteum collected from cows in the first 2 months of pregnancy24. Moreover, GAPDH and HPRT1 (along with SDHA) belonged to the least stably expressed RGs in bovine ovaries12. Nonetheless, GAPDH continues to be used as the internal control in many gene expression studies of bovine reproductive tissues32–35.
As transcriptional regulation of the establishment and maintenance of pregnancy is still poorly understood, our findings have important implications for future studies. The literature shows that RGs selected in some tissues under certain conditions are not necessarily the best performing genes under other circumstances. Based on the results of this study, we propose RPS9 and SUZ12 to be used as the best candidates for internal controls in expression experiments carried out on bovine caruncular epithelial cell cultures exposed to pregnancy-associated hormones, like P4, E2 and PGF2α.