The aim of the present pilot study was to determine the expression of RUNX2 gene in cumulus cells of women undergoing a certain protocol of assisted reproduction and to investigate its possible correlation with the controlled ovarian stimulation and pregnancy outcome. RUNX2 is known to be an essential factor for chondrocyte and osteoblast differentiation and bone formation. However, the expression of this gene in non- skeletal tissues and cells, as reported herein, may indicate other important functions of RUNX2. Specifically, in a recent study, RUNX2 gene expression was detected in periovulatory granulosa cells of rat and human ovaries, while a significant up-regulation of RUNX2 gene expression within 12 h after pre-treatment with hCG was revealed . Results obtained from our study demonstrated RUNX2 gene expression in cumulus cells of 13 out of 41 women that participated in the ART program with expression ratios ranging from 0.52 to 1.00. One possible interpretation of such RUNX2 mRNA levels could be the hCG administration performed in the present ART protocol, a finding that agrees with the previously reported results of Park and her co-workers, since hCG induced up-regulation of RUNX2 in rat preovulatory granulosa cells . On the other hand, it is of great interest that those women did not proceed to pregnancy, except for one woman with a weak RUNX2 gene expression that achieved pregnancy. Although hCG administration generally increases RUNX2 expression, RUNX2 absence in the other 28 women is remarkable. This finding could be attributed to the combination of hCG and rLH/rFSH administration, which may lead to inhibitory activities in the ovulation process of those women lacking RUNX2 gene expression. In the previously mentioned study of Park , rats were treated with pregnant mare serum gonadotropin (PMSG) and hCG to induce follicular development and ovulation, respectively, whereas in our study women were pre-treated with rLH and rFSH and administered with hCG. Moreover, in our study RUNX2 expression was studied at a specific time of the ART protocol, not having information about the expression profile of RUNX2 the previous days of the ovulation induction.
Nevertheless, it should be noted that a generally high expression profile of RUNX2 gene has been detected in periovulatory follicles, newly forming corpora lutea and corpora lutea from previous estrous cycles in rat ovary . In addition, a previous study revealed that the LH-surge induced RUNX2 expression is functionally linked to various aspects of luteal development by regulating the expression of specific luteal genes . Recent data outlined the antagonistic role of RUNX2 in regulating periovulatory gene expression. More specifically, RUNX2 up-regulates the expression of luteal genes, such asRgc32, Mmp13, Ptgds, Fabp6 and Abcb1a, whereas down-regulates the transcription of specific ovulatory genes in luteinizing granulosa cells [12, 17].
An interesting observation in our study concerns the higher number of follicles and retrieved oocytes in the group of women without RUNX2 gene expression, as well as higher serum LH levels. These findings suggest a possible inhibitory mechanism of RUNX2 in different pathways involved in oocyte maturation. It is well documented that RUNX2, as a transcription regulator of cell differentiation and proliferation, is involved in many biochemical pathways. A recent study revealed the transcriptional induction of hyaluronan and proteoglycan link protein 1 (Hapln1) gene expression by RUNX2 activity, through endogenous RUNX2 binding to the Hapln1 promoter region . It should be mentioned that Hapln1 enhances COC expansion by acting as a stabilizer of the cumulus matrix, thus promoting a successful ovulation . However, the present data correlate RUNX2 expression in cumulus cells with a lower number of follicles and retrieved oocytes. It could be suggested that the interaction Hapln1-RUNX2 is time-specific. RUNX2 expression was studied on the day of oocyte collection. Probably a daily study of expression after hCG administration and before oocyte collection may reveal that RUNX2-HalpnI is expressed until cumulus cells matrix is expanded and then declines acting as an activation mechanism for the rest of the procedure. Furthermore, in rat ovary Hapln1 expression was affected by both RUNX1 and RUNX2 expression , indicating the need for further investigation of RUNX2-RUNX1 interaction in human cumulus cells. Interestingly, RUNX1 transcription was found to be suppressed directly by RUNX2 over-expression in preovulatory granulosa cells, providing another insight of RUNX genes involvement in ovulation .
Regarding E2 levels at the day of hCG administration, several studies in the literature confirm that E2 levels play a fundamental role in cytoplasmic maturity, as well as in the quality of the embryos . The way estrogens affect RUNX2 activity as well as RUNX2-ER-α interactions are both complex. ER-α forms dimers with RUNX2 and the stimulatory effect of estrogens on RUNX2 activity is lost when the DNA binding domain of the estrogen receptor is eliminated. In cultured osteoblasts estrogen enhances RUNX2 activity in dose and estrogen-receptor dependent ways without changes in total RUNX2 levels or its affinity for DNA. On the other hand, in COS7 cells E2-bound ER-α suppresses the transactivation activity of RUNX2. Furthermore, E2 inhibits RUNX2 in late osteoblast cultures, while in early cultures RUNX2 may be stimulated by E2 . ER-α interacts with RUNX2 through multiple domains and the repression occurs independently of the activation of estrogen response element containing genes. RUNX2 suppresses estrogen activity by decreasing the effect of estradiol on reporter gene expression driven by the estrogen receptor response element . Recent studies on breast cancer cells disclosed that ER-α physically binds RUNX2 and inhibits expression of several RUNX2 target genes, providing a strong antagonistic correlation between the two genes in a different cellular type [22, 23]. This opposing effect is verified by our results. The non-parametric correlation between RUNX2 and E2 is statistically significant and indicates negative association (r = −0.40, p = 0.010). Comparing E2 levels between women with and without RUNX2 expression, women without expression present higher levels to a statistically significant degree (2957 pg/ml versus 1874 pg/ml, p = 0.013). An explanation of higher E2 levels in the group without RUNX2 expression in our study could be the down–regulation phenomenon in RUNX2 expression that may occur in this group of patients. Since E2-bound ER-α suppresses RUNX2 in a strong and specific mechanism, additional changes in ER-α expression between groups, apart from the different E2 levels, could also possibly affect RUNX2 expression, although ER-α has not been investigated in our study.