ESTROGEN- GROWTH HORMONE INTERACTION IN BONE CELLS OF THE OSTEOGENIC LINEAGE: GROWTH HORMONE ANABOLIC ACTIVITY ON HUMAN OSTEOBLASTS AND THEIR MESENCHYMAL PRECURSORS IS MODULATED BY 17ΒETA-ESTRADIOL THROUGH A POST RECEPTOR MECHANISM

In aging the loss of bone mass goes along with a decline of 17beta-estradiol (E2) and Growth Hormone (GH), which are known to be bone anabolic factors. Recent evidence demonstrated in several cell lines an interplay between E2 and GH at a post receptor level. I thus evaluated the possible cross-talk between these two hormones in human osteoblast cells (hOBs) in primary culture and in their mesenchymal precursors (hMSCs). E2 (10-8M) given 60 min before GH (5ng/ml) enhanced both GH intracellular pathway in hOBs and the transcription of the evaluated GH target genes involved in osteoblast activity and matrix deposition, osteopontin (OPN), bone sialoprotein (BSP) and insulin like growth factor 2 (IGF2). E2 effects occurred by decreasing the protein levels of SOCS2, one of the main GH signaling inhibitors, through an increase in SOCS2 ubiquitination and consequent degradation. This effect was blunted by pre-treating the cells with the proteasome inhibitor MG132 (5µM). Interestingly this effect did not involve an E2 mediated genomic activity as Actinomycin D (5µM) pre-treatment did not prevent E2 modulation of SOCS2 levels (Bolamperti S. et al., 2013). Further experiments demonstrated that this short term effect on SOCS2 levels was maintained over time: after 3h of E2 treatment there was still a decrease in SOCS2 levels. Yet, at this time point, the effect occurred via an inhibition of the transcription of SOCS2 gene. The fact that E2 negative regulation of the GH inhibitor SOCS2 involves an initially rapid protein degradation maintained for a longer time by a decrease in its gene expression strengthens the importance and the physiological relevance of this modulation for osteoblast activity. I was therefore interested to investigate whether or not two SERMs often used in clinics, Tamoxifen (Tam) and Raloxifene (Ral), share the same effect on GH signaling as E2. Cells treated with Tam (10-10M) or Ral (10-8M) 60 min before GH showed a trend to increase STAT5 phosphorylation even though reaching statistical significance, despite an observed reduction of SOCS2 levels with the SERMs alone. After 3h treatment no modulation of SOCS2 transcription was detected with either Tam, or Ral. These data suggest that despite their general estrogen agonistic properties on bone, none of the two drugs displayed the same features as E2. The combined effect of E2 and GH was evaluated also in mesenchymal stem cells (MSCs) obtained from human bones. The cells were first tested for plastic adherence, for differentiation capability towards adipocytes or osteoblasts, and for the positivity to CD73, CD105, and CD90 following an NIH protocol. In these osteoblast precursors, pretreatment of E2 60 min before GH, increased STAT5 phosphorylation induced by GH and decreased SOCS2 levels, as shown in hOBs. Considering the lack of information about GH action in stromal precursors, we evaluated GH action in the isolated hMSCs, focusing on its possible role in both osteogenesis and adipogenesis. The results showed that long term GH treatment (5ng/ml, 14 days, 3 times/week) increased early osteogenic genes and prevented adipogenesis in the isolated hMSCs. Given the important role of microRNAs in MSC commitment to osteoblastogenesis or adipogenesis, I analized if GH was able to increase miR-22 and miR-29c, considered amongst the main regulators of osteoblastogenesis or decrease miR-204, which is a regulator of adipogenesis. GH was able to upregulate the transcription of miR-22 and miR-29c, without affecting miR-204. E2 per se inhibited their transcription and, in the combined treatment with GH, E2 pre-treatment was able to inhibit the stimulatory effect of GH on miR22 and miR29c. In conclusion the study has shown a relevant hormone to hormone interaction; hence E2 can locally modulate GH activity potentiating its cellular signaling in osteoblasts and in their stromal precursors. In osteoblasts the increase in the activity of the GH signal transducers reflects an increase in the transcription of GH responsive genes involved in the regulation of the deposition and the turnover of minerals and in the control of osteoblasts and osteoclasts metabolism (Gehron Robey and Boskey, 2006). In mesenchymal stromal cells the E2 potentiating effect of GH signaling was not accompanied by a positive effect on miRNA expressions. This could be due to the strong inhibitory effect that E2 per se exerts on the evaluated miRNAs. Thus it can be suggested that the combined treatment of hMSCs with E2 and GH involves different mechanisms of the two hormones on microRNAs. The data from the present study suggest that circulating estrogen levels should be considered in the management of GH replacement therapy in GH deficiency. Since the E2-GH cross talk is not shared by SERMs, the positive effect of E2 on GH signaling should be taken into account while developing new estrogen receptor modulators molecules.