Apaf-1 is a transcriptional target for E2F and p53.

Loss of function of the retinoblastoma protein, pRB, leads to lack of differentiation, hyperproliferation and apoptosis. Inactivation of pRB results in deregulated E2F activity, which in turn induces entry to S-phase and apoptosis. Induction of apoptosis by either the loss of pRB or the deregulation of E2F activity occurs via both p53-dependent and p53-independent mechanisms. The mechanism by which E2F induces apoptosis is still unclear. Here we show that E2F1 directly regulates the expression of Apaf-1, the gene for apoptosis protease-activating factor 1. These results provide a direct link between the deregulation of the pRB pathway and apoptosis. Furthermore, because the pRB pathway is functionally inactivated in most cancers, the identification of Apaf-1 as a transcriptional target for E2F might explain the increased sensitivity of tumour cells to chemotherapy. We also show that, independently of the pRB pathway, Apaf-1 is a direct transcriptional target of p53, suggesting that p53 might sensitize cells to apoptosis by increasing Apaf-1 levels.

E2Fs regulate the expression of genes involved in differentiation, development, proliferation, and apoptosis.

The retinoblastoma protein (pRB) and its two relatives, p107 and p130, regulate development and cell proliferation in part by inhibiting the activity of E2F-regulated promoters. We have used high-density oligonucleotide arrays to identify genes in which expression changed in response to activation of E2F1, E2F2, and E2F3. We show that the E2Fs control the expression of several genes that are involved in cell proliferation. We also show that the E2Fs regulate a number of genes involved in apoptosis, differentiation, and development. These results provide possible genetic explanations to the variety of phenotypes observed as a consequence of a deregulated pRB/E2F pathway.

CDC25A phosphatase is a target of E2F and is required for efficient E2F-induced S phase.

Functional inactivation of the pRB pathway is a very frequent event in human cancer, resulting in deregulated activity of the E2F transcription factors. To understand the functional role of the E2Fs in cell proliferation, we have developed cell lines expressing E2F-1, E2F-2, and E2F-3 fused to the estrogen receptor ligand binding domain (ER). In this study, we demonstrated that activation of all three E2Fs could relieve the mitogen requirement for entry into S phase in Rat1 fibroblasts and that E2F activity leads to a shortening of the G(0)-G(1) phase of the cell cycle by 6 to 7 h. In contrast to the current assumption that E2F-1 is the only E2F capable of inducing apoptosis, we showed that deregulated E2F-2 and E2F-3 activities also result in apoptosis. Using the ERE2F-expressing cell lines, we demonstrated that several genes containing E2F DNA binding sites are efficiently induced by the E2Fs in the absence of protein synthesis. Furthermore, CDC25A is defined as a novel E2F target whose expression can be directly regulated by E2F-1. Data showing that CDC25A is an essential target for E2F-1, since its activity is required for efficient induction of S phase by E2F-1, are provided. Finally, our results show that expression of two E2F target genes, namely CDC25A and cyclin E, is sufficient to induce entry into S phase in quiescent fibroblasts. Taken together, our results provide an important step in defining how E2F activity leads to deregulated proliferation.

Induction of S-phase entry by E2F transcription factors depends on their nuclear localization.

The E2F transcription factors are essential for regulating the correct timing of activation of several genes whose products are implicated in cell proliferation and DNA replication. The E2Fs are targets for negative regulation by the retinoblastoma protein family, which includes pRB, p107, and p130, and they are in a pathway that is frequently found altered in human cancers. There are five members of the E2F family, and they can be divided into two functional subgroups. Whereas, upon overexpression, E2F-1, -2, and -3 induce S phase in quiescent fibroblasts and override G1 arrests mediated by the p16INK4A tumor suppressor protein or neutralizing antibodies to cyclin D1, E2F-4 and -5 do not. Using E2F-1 and E2F-4 as representatives of the two subgroups, we showed here, by constructing a set of chimeric proteins, that the amino terminus of E2F-1 is sufficient to confer S-phase-inducing potential as well as the ability to efficiently transactivate an E2F-responsive promoter to E2F-4. We found that the E2F-1 amino terminus directs chimeric proteins to the nucleus. Surprisingly, a short nuclear localization signal derived from simian virus 40 large T antigen could perfectly substitute for the presence of the E2F-1 amino terminus in these assays. Thus, nuclearly localized E2F-4, when overexpressed, displayed biological activities similar to those of E2F-1. Furthermore, we showed that nuclear localization of endogenous E2F-4 is cell cycle regulated, with E2F-4 being nuclear in the G0 and early G1 phases and mainly cytoplasmic after the pRB family members have become phosphorylated. We propose a novel mechanism for the regulation of E2F-dependent transcription in which E2F-4 regulates transcription only from G0 until mid- to late G1 phase whereas E2F-1 is active in late G1 and S phases, until it is inactivated by cyclin A-dependent kinase in late S phase.

Cyclin E-induced S phase without activation of the pRb/E2F pathway.

In cells of higher eukaryotes, cyclin D-dependent kinases Cdk4 and Cdk6 and, possibly, cyclin E-dependent Cdk2 positively regulate the G1- to S-phase transition, by phosphorylating the retinoblastoma protein (pRb), thereby releasing E2F transcription factors that control S-phase genes. Here we performed microinjection and transfection experiments using rat R12 fibroblasts, their derivatives conditionally overexpressing cyclins D1 or E, and human U-2-OS cells, to explore the action of G1 cyclins and the relationship of E2F and cyclin E in S-phase induction. We demonstrate that ectopic expression of cyclin E, but not cyclin D1, can override G1 arrest imposed by either the p16INK4a Cdk inhibitor specific for Cdk4 and Cdk6 or a novel phosphorylation-deficient mutant pRb. Several complementary approaches to assess E2F activation, including quantitative reporter assays in live cells, showed that the cyclin E-induced S phase and completion of the cell division cycle can occur in the absence of E2F-mediated transactivation. Together with the ability of cyclin E to overcome a G1 block induced by expression of dominant-negative mutant DP-1, a heterodimeric partner of E2Fs, these results provide evidence for a cyclin E-controlled S phase-promoting event in somatic cells downstream of or parallel to phosphorylation of pRb and independent of E2F activation. They furthermore indicate that a lack of E2F-mediated transactivation can be compensated by hyperactivation of this cyclin E-controlled event.

Post-transcriptional control regulates transforming growth factor alpha in the human carcinoma KB cell line.

Expression of epidermal growth factor receptor (EGF-R) antisense RNA results in a drastic reduction of EGF-R levels in the human carcinoma KB cell line and induces a reversion of their transformed phenotype (Moroni, M. C., Willingham, M. C., and Beguinot, L. (1992) J. Biol. Chem. 267, 2714-2722). We used parental and EGF-R antisense KB clones as a genetic system to study, in the same cell line, the role of transforming growth factor alpha (TGF-alpha) in the establishment and maintenance of the transformed phenotype. KB cells produce TGF-alpha mRNA, and their conditioned medium is able to sustain growth of antisense cells, mimicking the effect of exogenous EGF or TGF-alpha. In antisense cells there is a marked reduction of TGF-alpha mRNA steady-state levels. In addition, the decrease in TGF-alpha parallels the levels of residual EGF-R in the various antisense clones, indicating a direct correlation between receptors and growth factor levels. The addition of exogenous TGF-alpha (10 ng/ml) to antisense clones induces TGF-alpha levels. The half-life of TGF-alpha mRNA is 40-60 min in antisense cells and more than 8 h in parental KB cells, as determined by actinomycin D decay curves. This result indicates a predominant regulation of TGF-alpha mRNA at the post-transcriptional level. Nuclear run-on experiments show that there is only a marginal effect at the transcriptional level. We conclude that the autocrine loop responsible for the transformed phenotype of the human carcinoma KB cell line is dependent on both elevated levels of EGF-R and the presence of TGF-alpha. In addition, TGF-alpha is able to induce its own mRNA via a signal due to activation of the EGF-R acting predominantly at the post-transcriptional level.

In vitro and transgenic analysis of a human HOXD4 retinoid-responsive enhancer.

Expression of vertebrate Hox genes is regulated by retinoids in cell culture and in early embryonic development. We have identified a 185-bp retinoid-responsive transcriptional enhancer 5' of the human HOXD4 gene, which regulates inducibility of the gene in embryonal carcinoma cells through a pattern of DNA-protein interaction on at least two distinct elements. One of these elements contains a direct repeat mediating ligand-dependent interaction with retinoic acid receptors, and is necessary though not sufficient for the enhancer function. The HOXD4 enhancer directs expression of a lacZ reporter gene in the neural tube of transgenic mouse embryos in a time-regulated and regionally restricted fashion, reproducing part of the anterior neuroectodermal expression pattern of the endogenous Hoxd-4 gene. Administration of retinoic acid to developing embryos causes alterations in the spatial restriction of the transgene expression domain, indicating that the HOXD4 enhancer is also a retinoid-responsive element in vivo. The timing of the retinoic acid response differs from that seen with more 3' Hox genes, in that it occurs much later. This shows that the temporal window of competence in the ability to respond to retinoic acid differs between Hox genes and can be linked to specific enhancers. Mutations in the direct repeat or in a second element in the enhancer affect both retinoid response in culture and developmental regulation in embryos, suggesting that co-operative interactions between different factors mediate the enhancer activity. These data provide further support for a role of endogenous retinoids in regulation and spatial restriction of Hox gene expression in the central nervous system.

Retinoic acid induces stage-specific antero-posterior transformation of rostral central nervous system.

We report a time-course analysis of the effect of retinoic acid (RA) on the development of the mouse central nervous system (CNS) from the beginning of gastrulation throughout induction and patterning of the neural tube. RA administration induces three different, stage-specific alterations of brain development, indicating perturbation of different morphogenetic steps during the establishment of a neural pattern. In particular, treatment at mid-late streak stage (7.2-7.4 days post coitum (d.p.c.)) results in early repression of Otx2 expression in the posterior neuroectoderm of the head fold and in the ventral mid line, including the prechordal plate and the rostralmost endoderm, followed by loss of forebrain morphological and molecular identities, as revealed by analysis of the expression of regionally-restricted brain genes (Otx2, Otx1, Emx2, Emx1 and Dlx1). In these embryos, reduction of the Otx2 expression domain correlates with hindbrain expansion marked by rostral extension of the Hoxb-1 expression domain. Our analysis indicates that RA interferes with the correct definition of both planar and vertical morphogenetic signals at specific developmental stages by affecting gene expression in the regions which are likely either to produce or to respond to these signals. We suggest that retinoids may contribute to early definition of head from trunk structures by selecting different sets of regulatory genes.

Regulation of the human HOXD4 gene by retinoids.

Hox genes are developmentally regulated in mammalian embryogenesis, according to temporally and spatially restricted patterns which are affected by retinoids, vitamin A derivatives which have a function as, or at least mimic the action of, axis-specifying morphogens. In the human embryonal carcinoma cell line NT2/D1, HOX gene clusters are activated by at least two retinoids, all-trans- and 9-cis-retinoic acid (RA), in a 3' to 5' sequential cascade which reproduces the activation pattern observed in early embryogenesis. We have studied the regulation of the early activated HOXD4 gene, which is expressed in human embryogenesis in multiple transcripts generated by the developmentally controlled use of alternative transcription start sites and polyadenylation signals. Transfection of a 2.9 kb HOXD4 upstream genomic region linked to a reporter gene in NT2/D1 cells, allowed the identification of two different promoters and a distal enhancer element necessary for RA-dependent gene activation. This element confers to a heterologous promoter the ability to be induced by RA in NT2/D1 cells, and transactivated by alpha, beta and gamma retinoic acid receptors (RARs), but not retinoid X receptor (RXR), in COS-7 cells. DNase I footprinting analysis allowed the identification of four sequences which bind nuclear factors from both RA-induced NT2/D1 cells and embryonic tissues with similar patterns. The use of specific antibodies allowed the identification of at least RAR beta in some of the DNA-protein complexes, although the four sequences bind single RARs transfected in COS cells much less efficiently, or not at all, when compared to a canonical RAR responsive element. Induction of the HOXD4 promoter-enhancer in the presence of a selective RAR alpha antagonist indicated that the RAR alpha-dependent RAR beta activation is nevertheless a necessary step in HOX gene activation. Our results indicate that the ligand-dependent activity of RARs upon specific, cis-acting regulatory elements may have a key role in the induction of early activated HOX genes in response to retinoids. However, RARs represent only a fraction of the transcription factors interacting with the RA-responsive HOXD4 enhancer, which appears to be a complex element requiring specific combinations of nuclear factors for its proper function.

EGF-R antisense RNA blocks expression of the epidermal growth factor receptor and suppresses the transforming phenotype of a human carcinoma cell line.

We have used an antisense approach to investigate the role of overexpression of the normal human epidermal growth factor (EGF) receptor in the transformed phenotype of KB cells, which are a tumor derived human cell line. Initial experiments performed in vitro, showed that antisense RNA complementary to the entire coding region (AS-FL) or to parts of the EGF-R mRNA (AS-3', AS-5', and AS-K) effectively blocked translation of EGF-R mRNA. In addition, upon microinjection into KB cells, the in vitro synthesized antisense RNAs were able to inhibit transiently the synthesis of EGF-R. Inhibition was concentration-dependent, both in vitro and in cells, and the most effective constructs were those complementary to the entire coding region (AS-FL) or to the 3'-coding end of the mRNA (AS-3'). Transfection of the same EGF-R antisense RNA constructs into the human epidermoid carcinoma KB cell line gave rise to several clones stably expressing elevated levels of antisense RNA and resulting in low residual levels of EGF receptor. The most reduced clones exhibited a totally restored serum-dependent growth and were severely impaired in colony formation and growth in agar. In addition the severity of the phenotype was directly proportional to the residual amount of EGF-R expressed. We conclude that over-expression of normal EGF-R plays a direct primary role in the development of the transformed phenotype of this human cancer cell line.

Epidermal growth factor receptor in human brain tumors.

The expression of epidermal growth factor receptor (EGF-R) was examined in 27 primary human brain tumors (7 glioblastomas, 10 astrocytomas, 5 oligodendrogliomas, 1 schwannoma, 1 ganglioneuroma, 1 medulloblastoma, 1 ependymoma, 1 histiocytic lymphoma), in 6 brain metastases from lung carcinomas and in 20 meningiomas. Peritumoral tissues histologically normal excised surgically along with a large tumor were used as control. All plasma membranes from brain tissues tested showed specific EGF binding. The EGF receptor is expressed at low levels in the control human brain and at very high levels in 60% of the total intracranial tumors studied. When the various histological types of tumors were analyzed, the higher percentage of positive tumors was found with the meningiomas (85%) and the glioblastomas (71%), while the lower percentage of positivity was found with the oligodendrogliomas (40%) and the astrocytomas (30%). A good correlation between binding and total amount of EGF-R protein detected by Western Blot was also observed.

Cloning of a DNA region of Actinoplanes teichomyceticus conferring teicoplanin resistance.

Teicoplanin is a glycopeptide antibiotic, produced by Actinoplanes teichomyceticus, active against Gram positive bacteria and recently introduced into clinical practice. It blocks cell wall biosynthesis by inhibiting peptidoglycan polymerization. The mechanism(s) of resistance of the producer strains of this class of antibiotics have not yet been characterized. We have constructed a genomic bank of A. teichomyceticus in Streptomyces lividans. A clone from this bank, PTR168, was able to confer resistance to teicoplanin on its sensitive host. The restriction map of plasmid pTR168 and the hybridization pattern to A. teichomyceticus DNA were determined; we have also studied the mechanism of this resistance which seems correlated with a reduced binding of the antibiotic to the cell wall.