Potential hazards to embryo implantation: A human endometrial in vitro model to identify unwanted antigestagenic actions of chemicals.

Embryo implantation is a crucial step in human reproduction and depends on the timely development of a receptive endometrium. The human endometrium is unique among adult tissues due to its dynamic alterations during each menstrual cycle. It hosts the implantation process which is governed by progesterone, whereas 17ß-estradiol regulates the preceding proliferation of the endometrium. The receptors for both steroids are targets for drugs and endocrine disrupting chemicals. Chemicals with unwanted antigestagenic actions are potentially hazardous to embryo implantation since many pharmaceutical antiprogestins adversely affect endometrial receptivity. This risk can be addressed by human tissue-specific in vitro assays. As working basis we compiled data on chemicals interacting with the PR. In our experimental work, we developed a flexible in vitro model based on human endometrial Ishikawa cells. Effects of antiprogestin compounds on pre-selected target genes were characterized by sigmoidal concentration-response curves obtained by RT-qPCR. The estrogen sulfotransferase (SULT1E1) was identified as the most responsive target gene by microarray analysis. The agonistic effect of progesterone on SULT1E1 mRNA was concentration-dependently antagonized by RU486 (mifepristone) and ZK137316 and, with lower potency, by 4-nonylphenol, bisphenol A and apigenin. The negative control methyl acetoacetate showed no effect. The effects of progesterone and RU486 were confirmed on the protein level by Western blotting. We demonstrated proof of principle that our Ishikawa model is suitable to study quantitatively effects of antiprogestin-like chemicals on endometrial target genes in comparison to pharmaceutical reference compounds. This test is useful for hazard identification and may contribute to reduce animal studies.

Notch1 activation reduces proliferation in the multipotent hematopoietic progenitor cell line FDCP-mix through a p53-dependent pathway but Notch1 effects on myeloid and erythroid differentiation are independent of p53.

Signaling mediated by activation of the transmembrane receptor Notch influences cell-fate decisions, differentiation, proliferation, and cell survival. Activated Notch reduces proliferation by altering cell-cycle kinetics and promotes differentiation in hematopoietic progenitor cells. Here, we investigated if the G(1) arrest and differentiation induced by activated mNotch1 are dependent on tumor suppressor p53, a critical mediator of cellular growth arrest. Multipotent wild-type p53-expressing (p53(wt)) and p53-deficient (p53(null)) hematopoietic progenitor cell lines (FDCP-mix) carrying an inducible mNotch1 system were used to investigate the effects of proliferation and differentiation upon mNotch1 signaling. While activated Notch reduced proliferation of p53(wt)-cells, no change was observed in p53(null)-cells. Activated Notch upregulated the p53 target p21(cip/waf) in p53(wt)-cells, but not in p53(null)-cells. Induction of the p21(cip/waf) gene by activated Notch was mediated by increased binding of p53 to p53-binding sites in the p21(cip/waf) promoter and was independent of the canonical RBP-J binding site. Re-expression of p53(wt) in p53(null) cells restored the inhibition of proliferation by activated Notch. Thus, activated Notch inhibits proliferation of multipotent hematopoietic progenitor cells via a p53-dependent pathway. In contrast, myeloid and erythroid differentiation was similarly induced in p53(wt) and p53(null) cells. These data suggest that Notch signaling triggers two distinct pathways, a p53-dependent one leading to a block in proliferation and a p53-independent one promoting differentiation.

Preparation of nuclear matrices from cultured cells: subfractionation of nuclei in situ.

Analyses of the different structural systems of the nucleus and the proteins associated with them pose many problems. Because these systems are largely overlapping, in situ localization studies that preserve the in vivo location of proteins and cellular structures often are not satisfactory. In contrast, biochemical cell fractionation may provide artifactual results due to cross-contamination of extracts and structures. To overcome these problems, we have developed a method that combines biochemical cell fractionation and in situ localization and leads to the preparation of a residual cellular skeleton (nuclear matrix and cytoskeletal elements) from cultured cells. This method's main feature is that cell fractionation is performed in situ. Therefore, structures not solubilized in a particular extraction step remain attached to the substrate and retain their morphology. Before and after each extraction step they can be analyzed for the presence and location of the protein under study by using immunological or cytochemical techniques. Thereby the in vivo origin of a protein solubilized in a particular extraction step is determined. The solubilized protein then may be further characterized biochemically. In addition, to allow analyses of proteins associated with the residual cellular skeleton, we have developed conditions for its solubilization that do not interfere with enzymatic and immunological studies.

Interactions of mutant p53 with DNA: guilt by association.

Since the very early days of p53 research, the gain of oncogenic activities by some mutant p53 proteins had been suspected as an important factor contributing to cancer progression. Considerable progress towards understanding the biology of mutant p53 has been made during the last years, the quintessence being the realization that the impact of mutant p53 proteins on the transcriptome of a tumor cell is much more global than previously thought. The emerging role of mutant p53 proteins in coordinating oncogenic signaling and chromatin modifying activities reveals an until now unsuspected function of these proteins as important modifiers of the oncogenic transcriptional response. Notwithstanding the fact that the sequence-specific DNA binding activity of mutant p53 proteins is impaired, they are still able to associate with specific loci on DNA by utilizing different mechanisms. The ability to associate with DNA appears to be crucial for the master role of mutant p53 proteins in coordinating oncogenic transcriptional responses.

Regulation of cellular senescence by Rb2/p130.

Growth regulatory functions of Rb2/p130, which aim at a sustained arrest such as in quiescent or differentiated cells, qualify the protein also to act as a central regulator of growth arrest in cellular senescence. In this respect, Rb2/p130 functions are connected to signaling pathways induced by p53, which is a master regulator in cellular senescence. Here, we summarize the pathways, which specify pRb2/p130 to control this arrest program and distinguish its functions from those of pRb/p105.

Dissection of transcriptional and non-transcriptional p53 activities in the response to genotoxic stress.

Following genotoxic stress, p53 either rescues a damaged cell or promotes its elimination. The parameters determining a specific outcome of the p53 response are largely unknown. In mouse fibroblasts treated with different irradiation schemes, we monitored transcriptional and non-transcriptional p53 activities and identified determinants that initiate an anti- or a pro-apoptotic p53 response within the context of p53-independent stress signaling. The primary, transcription-mediated p53 response in these cells is anti-apoptotic, while induction of p53-dependent apoptosis requires an additional, transcription-independent p53 activity, provided by high intracellular levels of activated p53. High intracellular levels of p53 were selectively generated after apoptosis-inducing high-dose UV-irradiation, and correlated with a strongly delayed upregulation of Mdm2. Following high-dose UV-irradiation, p53 accumulated in the cytoplasm and led to activation of the pro-apoptotic protein Bax. As p53-dependent Bax-activation is transcription-independent, we postulated that certain transcription-deficient mutant p53 proteins might also exert this activity. Indeed we found an endogenous, transcription-inactive mutant p53 that upon genotoxic stress induced Bax-activation in vivo. Our results demonstrate the impact and in vivo relevance of non-transcriptional mechanisms for wild-type and mutant p53-mediated apoptosis.

Cooperation between p53 and p130(Rb2) in induction of cellular senescence.

To determine pathways cooperating with p53 in cellular senescence when the retinoblastoma protein (pRb)/p16INK4a pathway is defunct, we stably transfected the p16INK4a-negative C6 rat glioma cell line with a temperature-sensitive mutant p53. Activation of p53(Val-135) induces a switch in pocket protein expression from pRb and p107 to p130(Rb2) and stalls the cells in late G1, early S-phase at high levels of cyclin E. Maintenance of the arrest depends on the functions of p130(Rb2) repressing cyclin A. Inactivation of p53 in senescent cultures restores the pocket proteins to initial levels and initiates progression into S-phase, but the cells fail to resume proliferation, likely due to DNA damage becoming apparent in the arrest and activating apoptosis subsequent to the release from p53-dependent growth suppression. The data indicate that p53 can cooperate selectively with p130(Rb2) to induce cellular senescence, a pathway that may be relevant when the pRb/p16INK4a pathway is defunct.

Evidence for free and metabolically stable p53 protein in nuclear subfractions of simian virus 40-transformed cells.

To determine functional subcellular loci of p53, a cellular protein associated with cellular transformation, we analyzed the nucleoplasmic, chromatin, and nuclear matrix fractions from normal mouse 3T3 cells, from methylcholanthren-transformed mouse (MethA) cells, and from various simian virus 40 (SV40)-transformed cells for the presence of p53. In 3T3 and MethA cells, p53 was present in all nuclear subfractions, suggesting an association of p53 with different structural components of the nucleus. In 3T3 cells, p53 was rapidly turned over, whereas in MethA cells, p53 was metabolically stable. In SV40-transformed cells, p53 complexed to large tumor antigen (large T) was found in the nucleoplasmic and nuclear matrix fractions, as described previously (M. Staufenbiel and W. Deppert, Cell 33:173-181, 1983). In addition, however, metabolically stable p53 not complexed to large T (free p53) was also found in the chromatin and nuclear matrix fractions of these cells. This free p53 did not arise by dissociation of large T-p53 complexes, suggesting that stabilization of p53 in SV40-transformed cells can also occur by means other than formation of a complex with large T.

Modulation of p53 protein expression during cellular transformation with simian virus 40.

We analyzed the relation of metabolic stabilization of the p53 protein during cellular transformation by simian virus 40 (SV40) to (i) expression of the transformed phenotype and (ii) expression of the large tumor antigen (large T). Analysis of SV40-tsA28-mutant-transformed rat cells (tsA28.3 cells) showed that both p53 complexed to large T and free p53 (W. Deppert and M. Haug, Mol. Cell. Biol. 6:2233-2240, 1986) were metabolically stable when the cells were cultured at 32 degrees C and expressed large T and the transformed phenotype. At the nonpermissive temperature (39 degrees C), large-T expression is shut off in these cells and they revert to the normal phenotype. In such cells, p53 was metabolically unstable, like p53 in untransformed cells. To determine whether metabolic stabilization of p53 is directly controlled by large T, we next analyzed the metabolic stability of complexed and free p53 in SV40 abortively infected normal BALB/c mouse 3T3 cells. We found that neither p53 in complex with large T nor free p53 was metabolically stable. However, both forms of p53 were stabilized in SV40-transformed cells which had been developed in parallel from SV40 abortively infected cultures. Our results indicate that neither formation of a complex of p53 with large T nor large-T expression as such is sufficient for a significant metabolic stabilization of p53. Therefore, we suggest that metabolic stabilization of p53 during cellular transformation with SV40 is mediated by a cellular process and probably is the consequence of the large-T-induced transformed phenotype.

Membrane interactions of simian virus 40 large T-antigen: influence of protein sequences and fatty acid acylation.

To sort out possible influences of protein sequences and fatty acid acylation on the plasma membrane association of simian virus 40 large T-antigen, we have analyzed the membrane interactions of carboxy-terminal fragments of large T-antigen, encoded by the adenovirus type 2 (Ad2+)-simian virus 40 hybrid viruses Ad2+ND1 and Ad2+ND2. The 28,000 (28K)-molecular-weight protein of Ad2+ND1 as well as the 42K and 56K proteins of Ad2+ND2 associate preferentially with membranous structures and were found in association with the membrane system of the endoplasmic reticulum and with plasma membranes. Neither the endoplasmic reticulum membrane- nor the plasma membrane-associated 28K protein of Ad2+ND1 is fatty acid acylated. We, therefore, conclude that fatty acid acylation is not necessary for membrane association of this protein and suggest that an amino acid sequence in this protein is responsible for its membrane interaction. In contrast, the 42K and 56K proteins of Ad2+ND2 in plasma membrane fractions contain fatty acid. However, the interaction of these proteins with the plasma membrane differs from that of the 28K protein of Ad2+ND1: whereas the 28K protein of Ad2+ND1 interacts stably with Nonidet P-40-soluble constituents of the plasma membrane, the 42K and 56K proteins of Ad2+ND2 are tightly bound to the Nonidet P-40-insoluble plasma membrane lamina. Thus, an amino acid sequence in the amino-terminal region of the 28K protein confers membrane affinity to these proteins, whereas a region between the amino-terminal end of the 42K protein of Ad2+ND2 and the amino-terminal end of the 28K protein of Ad2+ND1 contains a reactive site for fatty acid acylation. This posttranslational modification correlates with the stable association of the 42K and 56K proteins with the plasma membrane lamina. We suggest that the same sequences also mediate the proper plasma membrane association of large T-antigen in simian virus 40-transformed cells.

Different regulation of the p53 core domain activities 3'-to-5' exonuclease and sequence-specific DNA binding.

In this study we further characterized the 3'-5' exonuclease activity intrinsic to wild-type p53. We showed that this activity, like sequence-specific DNA binding, is mediated by the p53 core domain. Truncation of the C-terminal 30 amino acids of the p53 molecule enhanced the p53 exonuclease activity by at least 10-fold, indicating that this activity, like sequence-specific DNA binding, is negatively regulated by the C-terminal basic regulatory domain of p53. However, treatments which activated sequence-specific DNA binding of p53, like binding of the monoclonal antibody PAb421, which recognizes a C-terminal epitope on p53, or a higher phosphorylation status, strongly inhibited the p53 exonuclease activity. This suggests that at least on full-length p53, sequence-specific DNA binding and exonuclease activities are subject to different and seemingly opposing regulatory mechanisms. Following up the recent discovery in our laboratory that p53 recognizes and binds with high affinity to three-stranded DNA substrates mimicking early recombination intermediates (C. Dudenhoeffer, G. Rohaly, K. Will, W. Deppert, and L. Wiesmueller, Mol. Cell. Biol. 18:5332-5342), we asked whether such substrates might be degraded by the p53 exonuclease. Addition of Mg2+ ions to the binding assay indeed started the p53 exonuclease and promoted rapid degradation of the bound, but not of the unbound, substrate, indicating that specifically recognized targets can be subjected to exonucleolytic degradation by p53 under defined conditions.

Specific mismatch recognition in heteroduplex intermediates by p53 suggests a role in fidelity control of homologous recombination.

We demonstrate that wild-type p53 inhibits homologous recombination. To analyze DNA substrate specificities in this process, we designed recombination experiments such that coinfection of simian virus 40 mutant pairs generated heteroduplexes with distinctly unpaired regions. DNA exchanges producing single C-T and A-G mismatches were inhibited four- to sixfold more effectively than DNA exchanges producing G-T and A-C single-base mispairings or unpaired regions of three base pairs comprising G-T/A-C mismatches. p53 bound specifically to three-stranded DNA substrates, mimicking early recombination intermediates. The KD values for the interactions of p53 with three-stranded substrates displaying differently paired and unpaired regions reflected the mismatch base specificities observed in recombination assays in a qualitative and quantitative manner. On the basis of these results, we would like to advance the hypothesis that p53, like classical mismatch repair factors, checks the fidelity of homologous recombination processes by specific mismatch recognition.

Two immunologically distinct human DNA polymerase alpha-primase subpopulations are involved in cellular DNA replication.

Metabolic labeling of primate cells revealed the existence of phosphorylated and hypophosphorylated DNA polymerase alpha-primase (Pol-Prim) populations that are distinguishable by monoclonal antibodies. Cell cycle studies showed that the hypophosphorylated form was found in a complex with PP2A and cyclin E-Cdk2 in G1, whereas the phosphorylated enzyme was associated with a cyclin A kinase in S and G2. Modification of Pol-Prim by PP2A and Cdks regulated the interaction with the simian virus 40 origin-binding protein large T antigen and thus initiation of DNA replication. Confocal microscopy demonstrated nuclear colocalization of hypophosphorylated Pol-Prim with MCM2 in S phase nuclei, but its presence preceded 5-bromo-2'-deoxyuridine (BrdU) incorporation. The phosphorylated replicase exclusively colocalized with the BrdU signal, but not with MCM2. Immunoprecipitation experiments proved that only hypophosphorylated Pol-Prim associated with MCM2. The data indicate that the hypophosphorylated enzyme initiates DNA replication at origins, and the phosphorylated form synthesizes the primers for the lagging strand of the replication fork.

Upregulation of mdm-2 expression in Meth A tumor cells tolerating wild-type p53.

Overexpression of mouse wild-type p53 (wt p53) in mouse Meth A tumor cells after transfection of wt p53 encoding vectors induced a strong growth-inhibitory response. Cells of only few of randomly selected surviving colonies contained and expressed the transfected wt p53 specific DNA. Despite expressing authentic wt p53, such cells (MethAp53wt) exhibited a similar phenotype as the parental Meth A cells. These cells overexpressed the mdm-2 (mouse double minute-2) gene, both at the RNA and at the protein level. Recently, the MDM-2 protein has been identified as a cellular target of p53, which can abolish its tumor suppressor activity. We, therefore, suggest that MDM-2 has mitigated the growth-inhibitory effect of wt p53 in MethAp53wt cells. Upregulation of mdm-2 expression in MethAp53wt cells was mediated by wt p53, as analysis of Meth A cells carrying a tsp53 (p53Val135) revealed a strict dependence of mdm-2 upregulation upon wt p53 expression. Our results propose that a balanced ratio of MDM-2 and p53 will allow cells to tolerate a limited expression of wt p53. This tolerance is not mediated by a direct inactivation of wt p53 via complex formation with MDM-2, as the majority of both MDM-2 and wt p53 in MethAp53wt cells was not complexed to each other.

p53 and mdm2 are expressed independently during cellular proliferation.

We analysed p53 expression during proliferation of serum stimulated Swiss mouse 3T3 cells and of concanavalin A stimulated mouse spleen lymphocytes and correlated it to rate of DNA synthesis and to expression of PCNA. We also analysed mdm2 gene expression, as rising p53 levels during proliferation might require MDM2 protein expression to functionally antagonize p53 mediated growth inhibition. p53 protein synthesis closely paralleled DNA synthesis and PCNA expression, suggesting a direct involvement of p53 in cellular DNA synthesis. mdm2 expression in 3T3 cells could not be correlated with p53 expression and DNA synthesis and was not detected at all in stimulated lymphocytes. We conclude that p53 and mdm2 expression during proliferation are not functionally related and that mdm2 expression is not required for proliferation.

Immortalization of BALB/c mouse embryo fibroblasts alters SV40 large T-antigen interactions with the tumor suppressor p53 and results in a reduced SV40 transformation-efficiency.

In order to analyse the immortalizing and transforming potential of simian virus 40 (SV40), we compared the transformation efficiencies of SV40 in primary and in established BALB/c mouse fibroblasts. Five independently isolated clones of freshly immortalized normal fibroblasts (FTE cells) were established from precrisis BALB/c mouse embryo fibroblasts (pMEF cells) according to the protocol for establishing 3T3 cells (Todaro & Green, 1963). These cells expressed a wild-type p53 and were indistinguishable in all parameters analysed from original 3T3 cells kept in our laboratory. Using abortive infection to control gene dosage, followed by selection of transformed cells by cloning in soft agar, SV40 was able to transform primary cells with a much higher efficiency than 3T3 or FTE cells. Analysis of this unexpected result revealed that the different transformation efficiencies of SV40 in primary and established cells correlated with an altered cellular response to SV40 infection regarding metabolic stabilization of p53 complexed to large T during abortive infection. Whereas p53 in pMEF cells became stabilized upon abortive infection with SV40, p53 in 3T3 and FTE cells remained unstable. Our results strongly favour the hypothesis that metabolic stabilization and the ensuing higher levels of p53 in abortively infected cells enhance the transforming competence of large T.

Phosphorylation of p53 in primary, immortalised and transformed Balb/c mouse cells.

To address the question whether phosphorylation of p53 might be functionally involved in its metabolic stabilisation and in cellular transformation processes, we have analysed the phosphorylation of the cellular protein p53 in normal and transformed cells of Balb/c mouse origin. Two-dimensional tryptic peptide maps of metabolically unstable p53 from normal Balb/c 3T3 cells and of metabolically stable p53 from 3T3 cells transformed by Simian virus 40 (SV3T3 cells) revealed no qualitative differences between their phosphorylation sites. Except for the unique lack of one or possibly two sites, the phosphopeptide map of p53 from cells transformed by the chemical carcinogen methylcholanthrene (MethA cells), expressing two different, metabolically stable mutant forms of p53, was identical to that of wild-type p53 from normal and SV40-transformed 3T3 cells. These results suggest that no direct relationship exists between phosphorylation of p53, its metabolic stabilisation, and cellular transformation processes. We have included in our analyses p53 from primary Balb/c mouse embryo fibroblast (MEF) cells and the immortalised MEFP27 cell line, established from primary cells after 27 passages. The phosphorylation sites of p53 from these cells were found to be identical to that of p53 from 3T3 cells. In addition, primary and established (immortalised) Balb/c mouse cells revealed no major differences in abundance of p53 mRNA and p53 protein, as well as stability of p53. These results indicate that immortalisation of normal cells involves neither gross alterations of p53 expression nor changes in specific phosphorylation of p53.

A novel mechanism for covalent attachment of fatty acid to SV40 large T antigen.

The plasma membrane associated subclass of simian virus 40 large T antigen is specifically acylated with palmitic acid in vivo. To further analyze possible biological functions of fatty acid acylation, we developed a target-bound cell free in vitro acylation assay, in which immunopurified large T, bound to protein A-sepharose, was incubated with [3H]fatty acid. In this assay, large T was efficiently labeled with [3H]palmitic acid, but not with [3H]myristic acid. Thus the specificity of the in vivo labeling was preserved in vitro, too. The specific acylation of large T in vitro seemed to occur by an autocatalytic reaction, since it was found to be independent of added acyltransferases and exogenous energy. The energy for this reaction must be provided by the large T molecule itself, probably by an energy-rich internal ester bond. Our results provide evidence for a novel mechanism for the covalent attachment of fatty acids to proteins, which might also operate in vivo.

Analysis of biological and biochemical parameters for chromatin and nuclear matrix association of SV40 large T antigen in transformed cells.

We analysed biological and biochemical parameters for the association of the simian virus 40 (SV40) large tumor antigen (large T) with the cellular chromatin and the nuclear matrix in SV40-transformed cells. Nuclear subclasses of large T were isolated by in situ cell fractionation (Staufenbiel & Deppert, 1983) and first analysed for possible biological functions in the maintenance of cellular transformation. Like large T in SV40 wild-type transformants, large T in SV40 tsA mutant (tsA58)-transformed cells, expressing a temperature-dependent phenotype, was present in all nuclear subfractions (nucleoplasm, chromatin and nuclear matrix), when cells were kept at the growth temperature permissive for the expression of the transformed phenotype (32 degrees C). When tsA mutant-transformed cells were shifted to the non-permissive growth temperature (39 degrees C), they reverted to the normal phenotype. Concomitantly, large T lost its ability to associate with the cellular chromatin and the nuclear matrix, indicating that an association of large T with these subcellular structures may be important for the maintenance of cellular transformation. We next analysed the DNA-binding properties (sequence-specific binding to the SV40 origin of replication, ORI) of the nuclear subclasses of SV40 wild-type and of SV40 mutant large T defective in SV40 ORI binding in order to determine the influence of sequence-specific DNA binding on the association of large T with the chromatin and the nuclear matrix. Our detailed analyses show distinct differences in the ability of the various nuclear subclasses of large T to bind to the SV40 ORI, but suggest that the association of large T with the chromatin and the nuclear matrix is mediated by protein-protein interactions rather than by sequence-specific DNA binding.

DNA binding properties of murine p53.

We analysed the in vitro binding of p53 from normal (3T3) and from chemically transformed (Meth A) Balb/c mouse cells to double-stranded (ds-) DNA and to single-stranded (ss-) DNA by DNA-cellulose chromatography. We confirm previous findings that p53 in cellular extracts exhibits ds-DNA-binding activity (Lane and Gannon, 1983). In addition, we demonstrate that such p53 also binds to ss-DNA. Analyses with immunopurified p53 protein provide evidence that this DNA-binding activity is intrinsic to p53. DNA binding of p53 could not be inhibited by a monoclonal antibody specific for the C-terminal region. An N-terminal deletion mutant of p53 (Rovinski et al., 1987) exhibited similar DNA-binding properties as wild-type p53, indicating that the N-terminus also is dispensable for DNA binding. We further show a close correlation between the DNA-binding activity of p53 from 3T3 cells and its association with nuclear substructures.