Trophoblast giant cells express NF-kappa B2 during early mouse development.

To investigate whether transcription factors of the NF-kappa B family could play a role in early mammalian development, we have analyzed the expression of nfkb1, nfkb2, c-Rel, RelA, RelB, and bcl-3 from 6.5- to 10.5-day mouse embryo implantation sites. Our study shows that nfkb2 mRNA and protein are specifically localized in trophoblast giant cells throughout the stages analyzed. Trophoblast giant cells obtained upon in vitro cultures of 7.5-day ectoplacental cones display NF-kappa B DNA-binding activity that is supershifted by the anti-NF-kappa B2 antibody. Trophoblast giant cells are embryo-derived cells that form an interface between embryonic and maternal tissues during early mouse development; they are involved in decidual remodeling and expansion of the embryonic cavity, placenta formation, and possibly avoidance of maternal immune response to the embryo. Our study suggests that NF-kappa B2 could play a role in the modulation of genes expressed in trophoblast giant cells during the course of early embryogenesis, and therefore be relevant for tissue remodeling and morphogenesis of placenta.

The PBX-regulating protein PREP1 is present in different PBX-complexed forms in mouse.

Human PREP1, a novel homeodomain protein of the TALE super-family, forms a stable DNA-binding complex with PBX proteins in solution, a ternary complex with PBX and HOXB1 on DNA, and is able to act as a co-activator in the transcription of PBX-HOXB1 activated promoters (Berthelsen, J., Zappavigna, V., Ferretti, E., Mavilio, F., Blasi, F. , 1998b. The novel homeoprotein Prep1 modulates Pbx-Hox protein cooperatity. EMBO J. 17, 1434-1445; Berthelsen, J., Zappavigna, V., Mavilio, F., Blasi, F., 1998c. Prep1, a novel functional partner of Pbx proteins. EMBO J. 17, 1423-1433). Here we demonstrate the presence of DNA-binding PREP1-PBX complexes also in murine cells. In vivo, PREP1 is a predominant partner of PBX proteins in various murine tissues. However, the choice of PBX family member associated with PREP1 is largely tissue-type specific. We report the cloning and expression domain of murine Prep1 gene. Murine PREP1 shares 100% identity with human PREP1 in the homeodomain and 95% similarity throughout the whole protein. In the adult mouse, PREP1 is expressed ubiquitously, with peaks in testis and thymus. We further demonstrate the presence of murine Prep1 mRNA and protein, and of different DNA-binding PREP1-PBX complexes, in mouse embryos from at least 9.5 days p.c. Moreover, we show that PREP1 is present in all embryonic tissues from at least 7.5-17.5 days p.c with a predominantly nuclear staining. PREP1 is able to super-activate the PBX-HOXB-1 autoregulated Hoxb-1 promoter, and we show that all three proteins, PREP1, PBX and HOXB-1, are present together in the mouse rhombomere 4 domain in vivo, compatible with a role of PREP1 as a regulator of PBX and HOXB-1 proteins activity during development.

Expression of matrix metalloproteinases during murine chorioallantoic placenta maturation.

A large body of experimental evidence supports the participation of two groups of extracellular proteases, matrix metalloproteinases (MMPs), and plasminogen activators/plasmin, in tissue remodeling in physiological and pathological invasion. In the late mouse placenta, several tissue remodeling and cell invasion processes take place. Spongiotrophoblast migration into maternal decidua, as well as decidual extracellular matrix remodeling require the coordinated action of extracellular proteolytic enzymes. Via Northern and in situ hybridization, we have analyzed the spatio-temporal expression patterns of members of the MMP family (stromelysin-3, gelatinases A and B), as well as their inhibitors TIMP-1, -2 and -3 in late murine placenta (days 10.5 to 18.5 of gestation). Gelatinase activity in placental extracts was assessed by substrate zymography. Gelatinase A and stromelysin-3 were found to be prominently expressed in decidual tissue; shortly after midpregnancy, the decidual expression patterns of gelatinase A and stromelysin-3 became overlapping with each other, as well as with the expression domain of TIMP-2. On the other hand, gelatinase B transcripts were expressed only by trophoblast giant cells at day 10.5, and were downregulated at later stages. TIMP-1 and TIMP-3 transcripts were detected in decidual periphery at day 10.5, while later the expression was restricted to the endometrial stroma and spongiotrophoblasts, respectively. The areas of stromelysin-3 expression were the same (giant trophoblasts) or adjacent (decidua) to those where urokinase (uPA) transcripts were detected, suggesting a possible cooperation between these proteinases in placental remodeling. We generated mice doubly deficient for stromelysin-3 and uPA, and report here that these mice are viable and fertile. Furthermore, these animals do not manifest obvious placental abnormalities, thereby suggesting the existence of compensatory/redundant mechanisms involving other proteolytic enzymes. Our findings document the participation of MMPs and their inhibitors in the process of late murine placenta maturation, and warrant the characterization of other members of the MMP family, like membrane type-MMPs, in this process.

Expression and function of the urokinase type plasminogen activator during mouse hemochorial placental development.

Mouse midlate placental development involves extensive tissue remodeling and cell invasion, processes which could be mediated by extracellular proteolytic enzymes. We have performed in situ expression analysis of urokinase type plasminogen activator (uPA), as well as functionally related molecules (uPA receptor, low density lipoprotein receptor-related protein, plasminogen activator inhibitor type-1) in day 10.5 to 18.5 post coitum (p.c.) murine placentas. In situ hybridization demonstrated the presence of uPA transcripts in the invasive trophoblast cells, in particular in glycogen-rich trophoblasts, a cell population that between embryonic days 12.5 and 15.5 infiltrates the maternal decidual tissue. In addition, we observed high uPA expression in the cells of uterine epithelium. Enzymatically active uPA was detected in both sites of uPA mRNA expression by in situ zymography. Expression and activity data suggest a role for this protease in the processes of cell invasion and uterine epithelial remodeling. Only low levels of uPA receptor (uPAR) transcripts were found in trophoblasts and decidual tissue at days 10.5 and 11.5 p.c. At the same stages, a prominent expression of plasminogen activator inhibitor type-1 (PAI-1) by spongiotrophoblasts and giant trophoblasts, as well as of LDL receptor-related protein (LRP) by spongiotrophoblasts and decidual cells could be detected, suggesting a role in regulating extracellular proteolysis in the area of fetomaternal interface. Analysis of uPA null placentas showed the presence of decidual extravascular fibrin deposits, which were not detected in wild type placentas. At the same time, the extent of infiltration of trophoblast cells in maternal decidual tissue, evaluated by anti-cytokeratin immunostaining, was similar in wild type and uPA null placentas. Our studies show that in murine hemochorial placentation, uPA has an essential role in the maintenance of the fibrinogenic/fibrinolytic balance in the decidua. The function of uPA in trophoblast invasion appears not to be indispensable, and its absence can be overcome by redundant or compensatory mechanisms.

Identification of a novel gene, PSD, adjacent to NFKB2/lyt-10, which contains Sec7 and pleckstrin-homology domains.

We have identified a novel human gene on chromosome 10q24 located contiguously to the 3' end of the NFKB2/lyt-10 gene in a tail to tail arrangement. We describe here a cDNA of 4307 bp, isolated from an adult human brain cDNA library, which contains an open reading frame encoding a putative protein of 645 amino acids with a predicted molecular weight of 71 kDa. Database homology searches indicate that this is a novel gene coding for a putative protein containing two discrete domains with significant homology to the Sec7 and pleckstrin-homology (PH) domains, respectively. We named this gene PSD (plekstrin-Sec7 domains gene). Northern blot analysis of a panel of RNAs from normal human tissues using the PSD cDNA as probe revealed the presence of three different tissue-specific transcripts of approximately 4.3, 2.3, and 1.8 kb, the longest of which is expressed only in brain. Our data suggest that the PSD gene may code for a protein related to a recently identified protein family containing both the Sec7 and the PH domains thought to be involved in signaling transduction processes.

Embryo implantation in mouse: fetomaternal coordination in the pattern of expression of uPA, uPAR, PAI-1 and alpha 2MR/LRP genes.

During the process of embryo implantation, trophoblast cells invade deep into uterine stroma and play a key role in establishing fetomaternal exchange of molecules. We have studied the in vivo expression patterns of the molecules of the urokinase system, during the process of mouse embryo implantation and early placentation. The sites of synthesis of urokinase-type plasminogen activator (uPA), uPA-receptor (uPAR), plasminogen activator inhibitor type 1 (PAI-1) and alpha 2-macroglobulin receptor/low density lipoprotein receptor-related protein (alpha 2MR/LRP) transcripts were determined by in situ hybridization. These genes were found to be expressed in a finely regulated pattern. High levels of uPA mRNA were found in invasive trophoblast cells, while the same cells did not appear to synthesize PAI-1. Starting from day 6.5, endothelial cells of newly forming vessels also transcribed uPA gene. uPAR and alpha 2MR/LRP were in all stages expressed by decidual tissue, and their expression domains overlapped in large areas. Immunohistochemistry with uPA and PAI-1 antibodies revealed areas of co-localization of these secreted proteins with the expression domains of uPAR and alpha 2MR/LRP, which is of great interest in view of the role of these two receptors in clearing uPA-PAI-1 complexes. In situ zymography demonstrated the presence of active uPA in the ectoplacental cone region at 7.5 and 8.5 days. Our studies outline the expression of a set of functionally related genes that is well coordinated between fetal and maternal tissues. This coordination may model other physiological and pathological invasive processes.

Cleavage of K-FGF produces a truncated molecule with increased biological activity and receptor binding affinity.

The K-FGF/HST (FGF-4) growth factor is a member of the FGF family which is efficiently secreted and contains a single N-linked glycosylation signal. To study the role of glycosylation in the secretion of K-FGF, we mutated the human K-fgf cDNA to eliminate the glycosylation signal and the mutated cDNA was cloned into a mammalian expression vector. Studies of immunoprecipitation from the conditioned medium of cells expressing this plasmid revealed that the lack of glycosylation did not impair secretion, however the unglycosylated protein was immediately cleaved into two NH2-terminally truncated peptides of 13 and 15 kD, which appeared to be more biologically active than the wild-type protein. These two proteins also showed higher heparin binding affinity than that of wt K-FGF. We have expressed in bacteria the larger of these two proteins (K140), in which the NH2-terminal 36 amino acids present in the mature form of K-FGF have been deleted. Mitogenicity assays on several cell lines showed that purified recombinant K140 had approximately five times higher biological activity than wild-type recombinant K-FGF. Studies of receptor binding showed that K140 had higher affinity than wt K-FGF for two of the four members of FGF receptor's family, specifically for FGFR-1 (flg) and FGFR-2 (bek). K140 also had increased heparin binding ability, but this property does not appear to be responsible for the increased affinity for FGF receptors. Thus removal of the NH2-terminal 36 amino acids from the mature K-FGF produces growth factor molecules with an altered conformation, resulting in higher heparin affinity, and more efficient binding to FGF receptors. Although it is not clear whether cleavage of K-FGF to generate K140 occurs in vivo, this could represent a novel mechanism of modulation of growth factor activity.

A retrovirus carrying the K-fgf oncogene induces diffuse meningeal tumors and soft-tissue fibrosarcomas.

The K-fgf/hst oncogene encodes a growth factor of the fibroblast growth factor (FGF) family and transforms cells through an autocrine mechanism which requires extracellular activation of its receptor(s). To identify the cell and tissue targets of K-fgf oncogenic potential in vivo, we constructed a recombinant retrovirus carrying the human K-fgf cDNA and injected it, together with helper Moloney murine leukemia virus, into immunocompetent as well as nude mice. The original construct was highly transforming in tissue culture but produced no detectable pathologies in vivo with the exception of a single fibrosarcoma which arose after a long latency. The virus produced by this tumor appears to have undergone a complex series of recombination events involving the helper Moloney murine leukemia virus. It encodes an Env/K-FGF fusion protein whose expression is under the control of a hybrid long terminal repeat. This virus (designated MFS, for meningeal fibrosarcoma) induces tumors in mice with high frequency and short latency. These neoplasms consist of aggressive fibrosarcomas of soft tissue as well as diffuse meningeal tumors originating from the dura mater that surround the whole central nervous system and cause severe hydrocephalus. The Env/K-FGF fusion protein expressed by the MFS virus has retained all of the biological properties of native K-FGF, including secretion, mitogenic activity, heparin binding, and neutralization by anti-K-FGF antibodies. These and other results indicate that the tumors induced by the MFS virus result from the oncogenic potential of K-FGF.

Angiogenic activity of the K-fgf/hst oncogene in neural transplants.

Using retrovirus-mediated gene transfer into neural transplants, we have expressed the human K-fgf/hst oncogene in the central nervous system. Single-cell suspensions of fetal rat brains were removed at embryonic days 13 and 14, exposed to a retroviral vector encoding the K-fgf oncogene and stereotaxically implanted into the caudate putamen of syngenic adult Fisher rats. Recipient animals were sacrificed at intervals of 6-16 months without evidence of neurological impairment. Mock-infected grafts showed the characteristic histopathological appearance of organotypically differentiated neural transplants. In contrast, grafts exposed to the K-fgf gene exhibited abundant capillary proliferation and capillary angiomas. By in situ hybridization analysis and immunohistochemistry, expression of K-fgf was detected in neural cells adjacent to vascular proliferations. Neurons and glia with abundant K-fgf transcripts were morphologically unaffected. In order to examine the transforming potential of the K-fgf gene in the nervous system, we combined retrovirus-mediated transfer of the K-fgf oncogene with a single transplacental exposure of the donor animals to the neurotropic carcinogen N-ethyl-N-nitrosourea (NEU). However, this combination of transforming agents did not result in tumor formation in the grafts. These results provide evidence for a powerful angiogenic effect of K-fgf on the developing brain in vivo.

Selective expression of high molecular weight basic fibroblast growth factor confers a unique phenotype to NIH 3T3 cells.

The phenotypes of NIH 3T3 cells transfected with basic fibroblast growth factor (bFGF) cDNAs that express only the high molecular weight (HMW) forms of bFGF, the 18-kDa form, or all forms were examined. Cells producing the 18 kDa or all forms of bFGF were transformed at high levels of growth factor expression but were nontransformed at low levels. Cell producing low levels of HMW forms of bFGF were growth impaired when compared with the parental cells. These cells tended to form multinucleated giant cells, did not grow in soft agar, were nontumorigenic, had a normal bFGF receptor number, and had a nontransformed morphology. Cells expressing high levels of HMW bFGFs had a transformed morphology and were tumorigenic. These data suggest a specific functional role for HMWbFGF.

The K-fgf/hst oncogene induces transformation through an autocrine mechanism that requires extracellular stimulation of the mitogenic pathway.

The K-fgf/hst oncogene encodes a secreted growth factor of the fibroblast growth factor (FGF) family. The ability of K-fgf-transformed cells to grow in soft agar and in serum-free medium is inhibited by anti-K-FGF neutralizing antibodies, consistent with an autocrine mechanism of transformation. The transformed properties of clones that express high levels of K-FGF are, however, only partially affected. To better define the autocrine mechanism of transformation by K-fgf and to determine whether receptor activation could occur intracellularly, we constructed two mutants of the K-fgf cDNA. Deletion of the sequences encoding the signal peptide suppressed K-fgf ability to induce foci in NIH 3T3 cells. A few morphologically transformed colonies were observed in cotransfection experiments, and they were found to express high levels of cytoplasmic K-FGF. However, their ability to grow in serum-free medium and in soft agar was inhibited by anti-K-FGF antibodies. Addition of a sequence encoding the KDEL endoplasmic reticulum and Golgi retention signal to the K-fgf cDNA led to accumulation of the growth factor in intracellular compartments. The ability of the KDEL mutant to induce foci in NIH 3T3 cells was much lower than that of the wild-type cDNA, and also in this case the transformed phenotype was reverted by anti-K-FGF antibodies. These and other findings indicate that the transformed phenotype of cells expressing a nonsecretory K-FGF is due to the extracellular activation of the receptor by the small amounts of growth factor that these cells still release. Thus, transformation by K-fgf appears to be due to an autocrine growth mechanisms that requires activation of the mitogenic pathway at the cell surface.

A murine fibroblast growth factor (FGF) receptor expressed in CHO cells is activated by basic FGF and Kaposi FGF.

We have cloned a murine cDNA encoding a tyrosine kinase receptor with about 90% similarity to the chicken fibroblast growth factor (FGF) receptor and the human fms-like gene (FLG) tyrosine kinase. This mouse receptor lacks 88 amino acids in the extracellular portion, leaving only two immunoglobulin-like domains compared to three in the chicken FGF receptor. The cDNA was cloned into an expression vector and transfected into receptor-negative CHO cells. We show that cells expressing the receptor can bind both basic FGF and Kaposi FGF. Although the receptor binds basic FGF with a 15- to 20-fold higher affinity, Kaposi FGF is able to induce down-regulation of the receptor to the same extent as basic FGF. The receptor is phosphorylated upon stimulation with both FGFs, DNA synthesis is stimulated, and a proliferative response is produced in cells expressing the receptor, whereas cells expressing the cDNA in the antisense orientation show none of these responses to basic FGF or Kaposi FGF. Thus this receptor can functionally interact with two growth factors of the FGF family.

Protection of mice against tumor growth by immunization with an oncogene-encoded growth factor.

The K-fgf/hst oncogene encodes a growth factor of the fibroblast growth factor (FGF) family that is secreted and transforms cells through a mechanism of autocrine cell proliferation. K-fgf-transformed cells are highly tumorigenic in immunocompetent allogeneic and syngeneic animals. BALB/c mice were immunized with a bacterial fusion protein consisting of a portion of the MS2 polymerase and of the human K-FGF precursor lacking only the first 4 amino acids or with a recombinant protein corresponding to the mature, secreted form of K-FGF (176 amino acids). They were then challenged with syngeneic K-fgf- or H-ras-transformed cells. Vaccinated animals exhibited a significant degree of protection against tumor induction, which was specific for K-fgf-transformed cells and correlated with the ability of the immunized mice to produce high titers of anti-K-FGF antibodies. Thus immunization with a single oncogene product can protect animals against tumor cells expressing this oncogene.

Transformation by basic fibroblast growth factor requires high levels of expression: comparison with transformation by hst/K-fgf.

Basic fibroblast growth factor is a potent mitogen for a variety of cell types and has been suggested to have transforming activity. To test this hypothesis, we have introduced a human bFGF cDNA into NIH 3T3 cells either by DNA transfection or by retrovirus infection. We have compared the properties of cell lines obtained with cells prepared similarly but expressing the hst/K-fgf growth factor. While bFGF does not contain an amino terminal signal sequence and is not secreted from cells in which it is synthesized, hst/K-fgf does contain a signal sequence and is secreted from cells. Our results show that the transformed phenotype correlates directly with the level of bFGF expression, since all transformed clones expressed high levels of bFGF, while nontransformed clones expressed comparatively low levels of bFGF. In contrast, even low levels of hst/K-fgf expression resulted in a transformed phenotype. These results suggest that bFGF is an inefficient transforming protein and that this may relate to its lack of secretion.

Overexpression of N-ras oncogene and epidermal growth factor receptor gene in human glioblastomas.

Five human glioblastoma cell lines were analyzed for oncogene activation with a panel of probes. Abnormal expression of the epidermal growth factor receptor (EGFr) gene was detected in four of five lines; N-ras oncogene overexpression was found in all five cell lines. These results were subsequently confirmed with fresh brain tumor and nonneoplastic brain tissue biopsy samples; increased expression of the N-ras proto-oncogene was observed in five of five glioblastomas, all of which also showed EGFr gene overexpression, but not in well-differentiated gliomas or in nonneoplastic brain tissue specimens. No significant differences in Ha-ras and Ki-ras expression were observed. Preliminary histochemical observations showed that intracellular levels of transforming growth factor alpha, a putative biochemical link between these two oncogenes, were significantly higher in glioblastoma cells than in controls.

Study of the c-Ha-ras-1 locus polymorphism in an Italian population with high incidence of gastric cancer.

The c-Ha-ras-1 proto-oncogene locus is characterized by a restriction fragment length polymorphism resulting from length variation in a variable tandem repetition (VTR) region downstream from the structural part of the gene. The presence of uncommon alleles at this region has been suggested to be an informative marker for the development of different malignancies, including solid tumours. In order to identify possible genetic markers of cancer risk, we studied the c-Ha-ras-1 locus polymorphism in an Italian population characterized by a high incidence of stomach tumours. Gastric cancer patients, some having first-degree relatives affected by the same malignancy, and control subjects were studied. A total of 176 DNAs was analysed by the Southern blotting technique with TaqI restriction enzyme. This yields a fragment containing the sequence of variable length (VTR) and also allows detection of a cleavage site polymorphism. Thirteen different alleles were detected and some new common and rare variants were found. Our results do not provide evidence that the inheritance of any allele may predispose to gastric malignancies. Segregation analysis carried out on 13 patients' families demonstrated, without exception, a Mendelian inheritance of patterns.

Gene dosage mutants at adenine phosphoribosyltransferase locus induced by colcemid in Chinese hamster V79-AP4 cells.

Pseudodiploid Chinese hamster V79-AP4 cells, functionally diploid at the adenine phosphoribosyltransferase (aprt) locus, were treated with colcemid, a well-known aneuploidizing agent, under various experimental conditions. Aneuploid and tetraploid cells and variants resistant to 10 micrograms/ml of 2,6-diaminopurine (DAP), which selects for presumptive aprt+/- heterozygotes in the untreated cells, were induced. Many of the induced variants were hypotetraploid with three (rather than four) chromosomes carrying the aprt gene. Dot-blot and Southern analysis of the DNA of these clones confirmed that they had three copies of the aprt gene. Their APRT specific enzymatic activity was 60-80% of that of wild-type V79-AP4. The results of these and other experiments suggest that in these variants resistance to DAP is due to an altered aprt gene dosage and point to a possible genetic effect of colcemid and other aneuploidizing agents in somatic mammalian cells.

A genetic analysis of the adenine phosphoribosyl transferase locus in Chinese hamster V79-AP4 cells: relevance to mutagenesis studies.

The chromosomal location of the autosomal locus aprt has been investigated in the permanent Chinese hamster cell line V79-AP4 by standard somatic cell genetics methodologies. Aprt is functionally dizygous in V79-AP4 and the 2 alleles map on 2 chromosome 3 homologs, in agreement with the chromosome assignment of the gene in Chinese hamster primary cells. Chromosome G-banding and a Southern blot analysis of V79-AP4 DNA, using as a probe the cloned Chinese hamster aprt gene, have not revealed any structural alteration at either of the 2 aprt alleles. One of the chromosomes 3 has, however, a terminal deletion in its long arm and is therefore morphologically marked. These findings could make V79-AP4 an interesting cell system for the study of mutational mechanisms at the aprt locus in Chinese hamster.