Retinoic acid-mediated down-regulation of Oct3/4 coincides with the loss of promoter occupancy in vivo

Oct3/4, a hallmark of the earliest stages of embryogenesis, is expressed in undifferentiated embryonal carcinoma (EC) and embryonic stem (ES) cells. Oct3/4 gene expression is dependent on the promoter region, the proximal enhancer and the newly identified distal enhancer. We have analysed in vivo occupancy of these elements. In undifferentiated EC and ES cells, strong footprints were detected at specific sites of all three regulatory elements. These were promptly lost upon RA treatment in ES cells and in P19 EC cells, in parallel with sharply reduced Oct3/4 mRNA levels. Thus, the occupancy of regulatory elements is coupled with Oct3/4 expression, and RA treatment causes coordinated factor displacement, leading to extinction of gene activity. In F9 EC cells, footprint was first abolished at the proximal enhancer. However, this loss of binding site occupancy did not result in a decrease in Oct3/4 mRNA levels. The partial factor displacement seen in F9 EC cells, combined with the observation that EC and ES cells utilize the proximal and distal enhancers in differential manner, indicate the complex pattern of Oct3/4 gene regulation, which could reflect a cell type- and lineage-specific expression of the gene in vivo.     

Regulation of E2F4 mitogenic activity during terminal differentiation by its heterodimerization partners for nuclear translocation

E2F/DP heterodimers play a pivotal role in the regulation of cell growth and differentiation. A decrease in E2F/DP activity occurs during cell cycle arrest and differentiation. However, very little is known about the specific role of the various E2F/DP members along the transition from proliferation to terminal differentiation. We have previously shown that E2F4 accounts for the vast majority of the endogenous E2F in differentiating muscle cells. Here, we show that E2F4, which lacks a nuclear localization signal (nls), is distributed in both the nucleus and the cytoplasm, in either asynchronously growing myoblasts or differentiated myotubes. E2F4 nuclear accumulation is induced by the binding in the cytoplasm with specific partners p107, pRb2/p130, and DP3delta, an nls-containing spliced form of DP3, which provide the nls. Although overexpression of E2F4/DP3delta reactivates the cell cycle in quiescent cells, the E2F4 nuclear accumulation induced by pRb2/p130 and p107 correlates with cell growth arrest Moreover, E2F4/DP3delta-induced cell cycle reactivation is efficiently counteracted by either p107 or pRb2/p130 overexpression. Reinduction in quiescent cells of DNA synthesis by E2F1/DP1 overexpression is abrogated by coexpression of pRb and is hampered by MyoD overexpression. Both pRb2/p130 and pRb, as well as MyoD, are up-regulated in myotubes. Accordingly, multinucleated myotubes, which are induced to reenter the S-phase by oncoviral proteins, are refractory to cell cycle reactivation by forced expression of E2F4/DP3delta or E2F1/DP1. Thus, E2F/DP repression represents only one of multiple redundant circuits that control the postmitotic state in terminally differentiated cells and that are targeted by adenovirus E1A and SV40 large T antigen.     

Inactivation of p27Kip1 by the viral E1A oncoprotein in TGFbeta-treated cells

The adenovirus oncoprotein E1A and the simian virus SV40 large T antigen can both reverse the strong growth-inhibitory effect of transforming growth factor(TGF)-beta on mink lung epithelial cells: exposure of TGF-beta causes these cells to arrest late in the G1 phase of the cell cycle (ref. 3). This arrest correlates with an increase in expression of the protein p15Ink4B (ref. 4), inactivation of the cyclin E/A-cdk2 complex by the inhibitory protein p27Kip1 (refs 5-7), and with the accumulation of unphosphorylated retinoblastoma protein. The rescue by E1A of cells from TGF-beta arrest is partly independent of its binding to retinoblastoma protein. Here we show that E1A directly affects the cyclin-dependent kinase inhibitor p27Kip1 in TGF-beta-treated cells by binding to it and blocking its inhibitory effect, thereby restoring the activity of the cyclin-cdk2 kinase complex. In this way, E1A can overcome the effect of TGF-beta and modulate the cell cycle. To our knowledge, E1A provides the first example of a viral oncoprotein that can disable a cellular protein whose function is to inhibit the activity of cyclin-dependent kinases.     

Different HPV16 E6/E7 oncogene expression patterns in epithelia reconstructed from HPV16-immortalized human endocervical cells and genital keratinocytes

Human papillomavirus type 16 (HPV16) E6/E7 oncogenes immortalize two types of human genital epithelial cells in vitro, endocervical cells and ectocervical or foreskin keratinocytes. Epithelia reconstructed in in vivo nude mouse implants or in vitro organotypic raft cultures from immortalized endocervical cells form higher grade dysplasia than those from keratinocytes. Here, we compared viral E6/E7 mRNA expression in immortalized cell lines of the three cell types using implants, rafts and in situ hybridization assays. Endocervical cells expressed E6/E7 throughout their reconstructed epithelia. In contrast, oncogenes were limited to basal cells for keratinocyte lower grade dysplasias. To study the role of the HPV16 promoter/enhancer in this repression in the upper layers of keratinocyte epithelia, new cell lines were established by immortalization with E6/E7 controlled by the SV40 promoter. The oncogenes were shown to be controlled from the SV40 elements after immortalization. Nevertheless, E6/E7 in the two cell types had the same cell-specific expression pattern as that controlled from the homologous HPV16 promoter. In addition, naturally occurring premalignant lesions having integrated HPV16 DNA expressed E6/E7 extensively in the high-grade dysplastic region of undifferentiated metaplasia. On the other hand, oncogene expression was restricted to lower layers in the lower grade dysplastic region of more mature differentiation. Our data suggest that keratinocytes have an inherent HPV16 promoter-nonspecific mechanism of repression. Apparently this mechanism, which can be acquired during maturation, is initially nonfunctional in in vitro and in vivo epithelia derived from metaplastic endocervical cells.     

Two mechanisms in the action of repressor deltaEF1: binding site competition with an activator and active repression

BACKGROUND: Counteraction between activators and repressors is crucial for the regulation of a number of cell-specific enhancers, where an activator and a repressor are mutually competitive in binding to the same site. DeltaEF1 is a repressor protein of delta1-crystallin minimal enhancer DC5 binding at the CACCT site, and inhibits activator deltaEF3 from binding to the overlapped site. It has two zinc finger clusters N-fin and C-fin, close to N- and C-termini, respectively, and a homeodomain in the middle. deltaEF1 also binds to the E2-box sequence CACCTG, and represses E2-box-dependent enhancers. RESULTS: The mechanism of the repressor action of deltaEF1 was investigated by examining various deletion mutants of deltaEF1 for their activity to repress delta1-crystallin enhancer fragment HN which contained DC5 sequence and an additional activator site. Both zinc finger clusters were found to be essential for DNA binding and repression, but the homeodomain was not. In addition, the NR domain close to the N-terminus was required for full repression. The NR domain showed active repression when fused to the Gal4 DNA binding domain. Active repression by deltaEF1, dependent on the NR domain, was also demonstrated in a situation where the binding sites of deltaEF1 and deltaEF3 were separated. N-fin and C-fin in their isolated forms bind the 5'-(T/C)ACCTG-3' and 5'-(t/C)ACCT-3' sequences, respectively, while the homeodomain showed no DNA binding activity. An analysis of DNA binding of the delta(Int)F form, having both N-fin and C-fin, indicated that a single DNA binding domain is assembled from two zinc finger clusters. CONCLUSION: Two mechanisms are involved in the repressor action of deltaEF1. First, a binding site competition with an activator which depends on the integrity of both zinc finger clusters, and second, an active repression to silence an enhancer which is attributed to the NR domain.     

Subcellular organization of the placenta in the Atlantic sharpnose shark, Rhizoprionodon terraenovae

Oncogenic transformation by human adenoviruses requires early regions 1A and 1B (E1A and E1B) and provides a model of multistep carcinogenesis. This study shows that the metabolic stabilization of p53 observed in adenovirus 5 (Ad5)-transformed cells can occur in untransformed cells expressing E1A alone. Stabilized p53 was localized to the nucleus and was indistinguishable from wild-type p53 with respect to its interactions with hsc70, PAb420, Ad5 p55E1B, and SV40 large T antigen. Moreover, binding of Ad5 p55E1B or SV40 large T antigen had no additional effect on p53 levels or turnover. Higher levels of p53 were also induced in a variety of cell types within 40 hr after transferring E1A genes. E1A also caused cells to lose viability by a process resembling apoptosis. The apoptosis appeared to involve p53, because p53 levels reverted to normal in surviving cells that had lost E1A, and E1B protected cells from the toxic effects of E1A. These results suggest that (1) the involvement of p53 in tumor suppression and/or apoptosis can be regulated at the level of protein turnover, and (2) a major oncogenic role for E1B is to counter cellular responses to E1A (i.e., stabilization of p53 and associated apoptosis) that preclude transformation by E1A alone. This represents the first physiological setting in which high levels of endogenous p53 are induced in response to an oncogenic challenge, with the apparent consequence of suppressing transformation.