Down-regulation of the tumor suppressor gene retinoic acid receptor beta2 through the phosphoinositide 3-kinase/Akt signaling pathway

The retinoic acid receptor beta2 (RARbeta2) is a potent, retinoid-inducible tumor suppressor gene, which is a critical molecular relay for retinoid actions in cells. Its down-regulation, or loss of expression, leads to resistance of cancer cells to retinoid treatment. Up to now, no primary mechanism underlying the repression of the RARbeta2 gene expression, hence affecting cellular retinoid sensitivity, has been identified. Here, we demonstrate that the phosphoinositide 3-kinase/Akt signaling pathway affects cellular retinoid sensitivity, by regulating corepressor recruitment to the RARbeta2 promoter. Through direct phosphorylation of the corepressor silencing mediator for retinoic and thyroid hormone receptors (SMRT), Akt stabilized RAR/SMRT interaction, leading to an increased tethering of SMRT to the RARbeta2 promoter, decreased histone acetylation, down-regulation of the RARbeta2 expression, and impaired cellular differentiation in response to retinoid. The phosphoinositide 3-kinase/Akt signaling pathway, an important modulator of cellular survival, has thus a direct impact on cellular retinoid sensitivity, and its deregulation may be the triggering event in retinoid resistance of cancer cells.     

Response of human rhabdomyosarcoma cell lines to retinoic acid: relationship with induction of differentiation and retinoic acid sensitivity

The ability of retinoids to induce growth inhibition associated with differentiation of diverse cell types makes them potent anti-cancer agents. We examined the effect of retinoic acid (RA) in cell lines derived from rhabdomyosarcoma (RMS), a malignant soft-tissue tumor committed to the myogenic lineage, but arrested prior to terminal differentiation. We showed that several RMS derived cell lines, including RD human rhabdomyosarcoma cells, are resistant to the growth-inhibitory and differentiation effects of RA. We established that this RA-resistance correlates with reduced expression and activity of RA-receptors in RD cells. We stably expressed either RARalpha, RARbeta, RARgamma, or RXRalpha expression vector into RD cells and found that only RARbeta or RARgamma induced a significant RA growth arrest without promoting differentiation indicating that changes in the amounts of RARs and RXRs are not sufficient to determine the RA myogenic response of rhabdomyosarcoma cells. Activation of RD cell differentiation by ectopic MRF4 expression enhanced RA-receptor activity and led to RA induction of differentiation. These studies demonstrate that RA-resistance of RD cells is linked to their lack of differentiation and suggest that the differentiation-promoting activity of RA requires factors other than RAR-RXR heterodimers.     

Retinoic acid via RARalpha inhibits the expression of 24-hydroxylase in human prostate stromal cells

25-Hydroxyvitamin D(3)-24-hydroxylase (24-hydroxylase) is an important inactivating enzyme and its expression is induced by 25-hydroxyvitamin D3 (25OHD3) and 1alpha,25-dihydroxyvitamin D3 (1alpha,25-(OH)2D3) through action of heterodimers of vitamin D receptor (VDR) and retinoid X receptor (RXR). RXRs also act as heterodimer partners for retinoic acid receptors (RARs), mediating the action of all-trans-retinoic acid (ATRA). Prostate stroma plays a crucial role in prostate cancer development and benign prostatic hyperplasia. We demonstrate here that ATRA markedly reduced the expression of 24-hydroxylase mRNA induced by 25OHD3 and 1alpha,25-(OH)2D3 in human prostatic stromal cells P29SN and P32S but not in epithelial cells PrEC or cancer cells LNCaP. By using transfection and RAR-selective ligands, we found that the inhibitory effect of ATRA on 24-hydroxylase expression in stromal cells was mediated by RARalpha but not by RARbeta. Moreover, the ATRA-induced expression of RARbeta was also mediated by RARalpha. The combined treatment of 1alpha,25-(OH)2D3 and RARalpha agonist Am80 at 10 nM exhibited strong growth-inhibitory effect whereas either alone had no effect. Our data suggest that ATRA suppresses 24-hydroxylase expression through RARalpha-dependent signaling pathway and can enhance vitamin D action in suppression of cell growth.     

Prepubertal testis development relies on retinoic acid but not rexinoid receptors in Sertoli cells

Sertoli cells (SC) are instrumental to stem spermatogonia differentiation, a process that critically depends on retinoic acid (RA). We show here that selective ablation of RA receptor alpha (RARalpha) gene in mouse SC, singly (Rara(Ser-/-) mutation) or in combination with RARbeta and RARgamma genes (Rara/b/g(Ser-/-) mutation), abolishes cyclical gene expression in these cells. It additionally induces testis degeneration and delays spermatogonial expression of Stra8, two hallmarks of RA deficiency. As identical defects are generated upon inactivation of RARalpha in the whole organism, our data demonstrate that all the functions exerted by RARalpha in male reproduction are Sertoli cell-autonomous. They further indicate that RARalpha is a master regulator of the cyclical activity of SC and controls paracrine pathways required for spermatogonia differentiation and germ cell survival. Most importantly, we show that the ablation of all RXR (alpha, beta and gamma isotypes) in SC does not recapitulate the phenotype generated upon ablation of all three RARs, thereby providing the first evidence that RARs exert functions in vivo independently of RXRs.     

Retinoic acid causes cell growth arrest and an increase in p27 in F9 wild type but not in F9 retinoic acid receptor beta2 knockout cells

We have previously shown that an F9 teratocarcinoma retinoic acid receptor beta(2) (RARbeta(2)) knockout cell line exhibits no growth arrest in response to all-trans-retinoic acid (RA), whereas F9 wild type (Wt), F9 RARalpha(-/-), and F9 RARgamma(-/-) cell lines do growth arrest in response to RA. To examine the role of RARbeta(2) in growth inhibition, we analyzed the cell cycle regulatory proteins affected by RA in F9 Wt and F9 RARbeta(2)(-/-) cells. Flow microfluorimetry analyses revealed that RA treatment of F9 Wt cells greatly increased the percentage of cells in the G1/G0 phase of the cell cycle. In contrast, RA did not alter the cell cycle distribution profile of RARbeta(2)(-/-) cells. In F9 Wt cells, cyclin D1, D3, and cyclin E protein levels decreased, while cyclin D2 and p27 levels increased after RA treatment. Compared to the F9 Wt cells, the F9 RARbeta(2)(-/-) cells exhibited lower levels of cyclins D1, D2, D3, and E in the absence of RA, but did not exhibit further changes in the levels of these cell cycle regulators after RA addition. Since RA significantly increased the level of p27 protein (approximately 24-fold) in F9 Wt as compared to the F9 RARbeta(2)(-/-) cells, we chose to study p27 in greater detail. The p27 mRNA level and the rate of p27 protein synthesis were increased in RA-treated F9 Wt cells, but not in F9 RARbeta(2)(-/-) cells. Moreover, RA increased the half-life of p27 protein in F9 Wt cells. Reduced expression of RARbeta(2) is associated with the process of carcinogenesis and RARbeta(2) can mediate the growth arrest induced by RA in a variety of cancer cells. Using both genetic and molecular approaches, we have identified some of the molecular mechanisms, such as the large elevation of p27, through which RARbeta(2) mediates these growth inhibitory effects of RA in F9 cells.     

Murine RARbeta4 displays reduced transactivation activity, lower affinity for retinoic acid, and no anti-AP1 activity

The biological actions of retinoic acid (RA) are mediated by retinoic acid receptors (RARalpha, beta, and gamma) and retinoid X receptors (RXR alpha, beta, and gamma). Each of the RARs is expressed as four to seven different isoforms. Four isoforms of RAR beta (beta1, beta2, beta3, and beta4), which differ only in their N-terminal sequence (A domain) have been described. These RARbeta isoforms display a specific pattern of expression in developing and adult animals and are highly evolutionarily conserved suggesting that they mediate distinct cellular effects of vitamin A. Experiments were performed to examine directly the RA-binding activity, transactivation activity, and anti-AP1 activity of each of these four RARbeta isoforms. The results demonstrate that RARbeta1, beta2, and beta3 bind RA with a similar K(d) value, have a similar EC(50) value in RA-dependent transactivation assays and inhibit AP1 activity to a similar level. By contrast, RARbeta4 has an elevated K(d) for RA, an increased EC(50) value in RA-dependent transactivation assays and does not display the ability to inhibit AP1 activity. This provides additional evidence that at least one RAR isoform, RARbeta4, may mediate distinct activities within a cell. Furthermore, these data suggest that the presence of an A domain in RARbeta is important for modulating these activities of RARs.