Matrix Gla Protein drives stemness and tumor initiation in ovarian cancer.

Ovarian cancer (OC) displays the highest mortality among gynecological tumors, mainly due to early peritoneal dissemination, the high frequency of tumor relapse following primary debulking, and the development of chemoresistance. All these events are thought to be initiated and sustained by a subpopulation of neoplastic cells, termed ovarian cancer stem cells (OCSC), that are endowed with self-renewing and tumor-initiating properties. This implies that interfering with OCSC function should offer novel therapeutic perspectives to defeat OC progression. To this aim, a better understanding of the molecular and functional makeup of OCSC in clinically relevant model systems is essential. We have profiled the transcriptome of OCSC vs. their bulk cell counterpart from a panel of patient-derived OC cell cultures. This revealed that Matrix Gla Protein (MGP), classically known as a calcification-preventing factor in cartilage and blood vessels, is markedly enriched in OCSC. Functional assays showed that MGP confers several stemness-associated traits to OC cells, including a transcriptional reprogramming. Patient-derived organotypic cultures pointed to the peritoneal microenvironment as a major inducer of MGP expression in OC cells. Furthermore, MGP was found to be necessary and sufficient for tumor initiation in OC mouse models, by shortening tumor latency and increasing dramatically the frequency of tumor-initiating cells. Mechanistically, MGP-driven OC stemness was mediated by the stimulation of Hedgehog signaling, in particular through the induction of the Hedgehog effector GLI1, thus highlighting a novel MGP/Hedgehog pathway axis in OCSC. Finally, MGP expression was found to correlate with poor prognosis in OC patients, and was increased in tumor tissue after chemotherapy, supporting the clinical relevance of our findings. Thus, MGP is a novel driver in OCSC pathophysiology, with a major role in stemness and in tumor initiation.

The interplay between NCAM and FGFR signalling underlies ovarian cancer progression.

Epithelial ovarian carcinoma (EOC) is an aggressive neoplasm, which has often disseminated to peritoneal cavity at the time of diagnosis. In order to better understand the molecular mechanisms behind EOC progression, we investigated the expression and functional role of neural cell adhesion molecule (NCAM) in this tumour type.

Response of bovine endothelial cells to FGF-2 and VEGF is dependent on their site of origin: Relevance to the regulation of angiogenesis.

Angiogenesis, the formation of new capillary blood vessels, occurs almost exclusively in the microcirculation. This process is controlled by the interaction between factors with positive and negative regulatory activity. In this study, we have compared the effect of two well described positive regulators, vascular endothelial growth factor (VEGF) and basic fibroblast growth factor (FGF-2) on bovine adrenal cortex-derived microvascular endothelial (BME) and bovine aortic endothelial (BAE) cells. The parameters we assessed included (a) cellular reorganization and lumen formation following exposure of the apical cell surface to a three-dimensional collagen gel; (b) organization of the actin cytoskeleton; (c) expression of thrombospondin-1 (TSP-1), an endogenous negative regulator of angiogenesis; and (d) extracellular proteolytic activity mediated by the plasminogen activator (PA)/plasmin system. We found that (a) collagen gel overlay induces rapid reorganization and lumen formation in BME but not BAE cells; (b) FGF-2 but not VEGF induced dramatic reorganization of actin microfilaments in BME cells, with neither cytokine affecting BAE cells; (c) FGF-2 decreased TSP-1 protein and mRNA expression in BME cells, an effect which was specific for FGF-2 and BME cells, since TSP-1 protein levels were unaffected by VEGF in BME cells, or by FGF-2 or VEGF in BAE cells; (d) FGF-2 induced urokinase-type PA (uPA) in BME and BAE cells, while VEGF induced uPA and tissue-type PA in BME cells with no effect on BAE cells. Taken together, these findings reveal endothelial cell-type specific responses to FGF-2 and VEGF, and point to the greater specificity of these cytokines for endothelial cells of the microvasculature than for large vessel (aortic) endothelial cells. Furthermore, when viewed in the context of our previous observation on the synergistic interaction between VEGF and FGF-2, our present findings provide evidence for complementary mechanisms which, when acting in concert, might account for the synergistic effect.

N-CAM modulates tumour-cell adhesion to matrix by inducing FGF-receptor signalling.

Loss of expression of neural cell-adhesion molecule (N-CAM) is implicated in the progression of tumour metastasis. Here we show that N-CAM modulates neurite outgrowth and matrix adhesion of beta-cells from pancreatic tumours by assembling a fibroblast-growth-factor receptor-4 (FGFR-4) signalling complex, which consists of N-cadherin, FGFR-4, phospholipase C gamma (PLC-gamma), the adaptor protein FRS2, pp60(c-src), cortactin and growth-associated protein-43 (GAP-43). Dominant-negative FGFR-4, inhibitors of FGFR signalling and anti-beta(1)-integrin antibodies repress matrix adhesion induced by N-CAM. FGF ligands can replace N-CAM in promoting matrix adhesion but not neurite outgrowth. The results indicate that N-CAM stimulates beta1-integrin-mediated cell-matrix adhesion by activating FGFR signalling. This is a potential mechanism for preventing the dissemination of metastatic tumour cells.

Cell adhesion in tumor invasion and metastasis: loss of the glue is not enough.

Tumor cells often show a decrease in cell-cell and/or cell-matrix adhesion. An increasing body of evidence indicates that this reduction in cell adhesion correlates with tumor invasion and metastasis. Two main groups of adhesion molecules, cadherins and CAMs, have been implicated in tumor malignancy. However, the specific role that these proteins play in the context of tumor progression remains to be elucidated. In this review, we discuss recent data pointing to a causal relationship between the loss of cell adhesion molecules and tumor progression. In addition, the direct involvement of these molecules in specific signal transduction pathways will be considered, with particular emphasis on the alterations of such pathways in transformed cells. Finally, we review recent observations on the molecular mechanisms underlying metastatic dissemination. In many cases, spreading of tumor cells from the primary site to distant organs has been characterized as an active process involving the loss of cell-cell adhesion and gain of invasive properties. On the other hand, various examples of metastases exhibiting a relatively benign (i.e. not invasive) phenotype have been reported. Together with our recent results on a mouse tumor model, these findings indicate that 'passive' metastatic dissemination can occur, in particular as a consequence of impaired cell-matrix adhesion and of tumor tissue disaggregation.

Characterization of novel clonal murine endothelial cell lines with an extended life span.

A murine endothelial cell line was recently established from microvessels that had invaded a subcutaneous sponge implant (Dong, Q. G.; Bernasconi, S.; Lostaglio, S., et al. Arterioscl. Thromb. Vasc. Biol. 17:1599-1604; 1997). From these sponge-induced endothelial (SIE) cells, we have isolated two subpopulations endowed with different phenotypic properties. Clone SIE-F consists of large, highly spread cells that have a relatively slow growth rate, form contact-inhibited monolayers, do not grow under anchorage-independent conditions, express elevated levels of thrombospondin-1 (TSP-1) and are not tumorigenic in vivo. In contrast, clone SIE-S2 consists of small, spindle-shaped cells that have a high proliferation rate, do not show contact-inhibition, grow under anchorage-independent conditions, express very low levels of TSP-1 and are tumorigenic in vivo. Both clones express the endothelial markers vascular endothelial-cadherin and vascular intercellular adhesion molecule-1, but do not express CD31 and E-selectin. In addition, SIE-S2 cells, but not SIE-F cells, express the alpha-smooth muscle actin isoform. SIE-S2 cells, but not SIE-F cells, are able to form branching tubes in fibrin gels. The SIE-F and SIE-S2 clones, which have properties of nontransformed and transformed cells, respectively, should provide useful tools to investigate physiological and pathological processes involving vascular endothelium.

Molecular mechanisms of tumor angiogenesis and tumor progression.

The formation of new blood vessels (angiogenesis) is crucial for the growth and persistence of primary solid tumors and their metastases. Furthermore, angiogenesis is also required for metastatic dissemination, since an increase in vascular density will allow easier access of tumor cells to the circulation. Induction of angiogenesis precedes the formation of malignant tumors, and increased vascularization seems to correlate with the invasive properties of tumors and thus with the malignant tumor phenotype. In the last few years, the discovery and characterization of tumor-derived angiogenesis modulators greatly contributed to our understanding of how tumors regulate angiogenesis. However, although angiogenesis appears to be a rate-limiting event in tumor growth and metastatic dissemination, a direct connection between the induction of angiogenesis and the progression to tumor malignancy is less well understood. In this review, we discuss the most recent observations concerning the modulation of angiogenesis and their implications in tumor progression, as well as their potential impact on cancer therapy.

Exogenous fibroblast growth factor-2 induces a transformed phenotype in vascular kaposi's sarcoma-like cells.

Vascular TTB cells derive from murine Kaposi's sarcoma-like dermal lesions and share several phenotypic features with AIDS-associated KS spindle cells. We have recently reported that fibroblast growth factor-2 (FGF-2) promotes dramatic cytoskeletal and morphological alterations in TTB cells, concomitant with the induction of an autocrine loop for hepatocyte growth factor and a relocalization of the urokinase receptor. Since all these alterations are hallmarks of cell transformation. we attempted to verify whether FGF-2 induces a transformed phenotype in TTB cells. Our results show that FGF-2-treated TTB cells acquire the ability to grow under anchorage-independent conditions. In addition, FGF-2 markedly reduced the levels of thrombospondin-1, an antiangiogenic and tumor suppressor protein, in TTB cells. Therefore, FGF-2 induces KS-like spindle cells to acquire properties characteristic of transformed cells. This suggests that FGF-2 plays a pathogenetic role in KS not only by promoting angiogenesis, but also by conferring a transformed phenotype upon KS cells. In light of previous reports on Tat-induced release of FGF-2 into the extracellular space, our findings may provide an additional mechanism for the observed synergism between Tat and FGF-2 in the pathogenesis of KS.

Pituitary cotransplantation significantly improves the performance, insulin content, and vascularization of renal subcapsular islet grafts.

Because the pituitary contains hormones with beta-cell trophic activity, we evaluated whether cotransplantation of pituitary tissue with pancreatic islets might be beneficial for islet graft function and survival. Streptozotocin diabetic nude mice were transplanted under the kidney capsule with 150 handpicked islets alone or mixed with two diced pituitaries and were then followed for 4 weeks. Mice transplanted with mixed islet/pituitary grafts had higher levels of circulating prolactin (PRL) than mice transplanted with islets only, while serum cortisol, growth hormone, and follicle-stimulating hormone were similar in the two groups. After transplantation, recipients of mixed islet/pituitary grafts showed a more pronounced decrease in glycemic levels and higher systemic insulin levels than mice transplanted only with islets. Mixed islet/pituitary grafts were macroscopically characterized by an excellent vascularization and were biochemically characterized by higher insulin and PRL content than pure islet grafts. Histologically, posttransplantation remodeling originated a hybrid organ in which healthy, well-vascularized islets were adjacent to pituitary cell clusters. Transplantations performed to address the specific effect of the anterior versus the intermediate pituitary lobes indicated the former as responsible for the improved function of cotransplanted islets. Mixed islet/pituitary grafts composed of anterior lobes were also the best vascularized and were histologically characterized by the presence of many folliculo-stellate cells. In conclusion, we obtained evidence that pituitary cotransplantation significantly improves the function, insulin content, and vascularization of suboptimal islet grafts. Evidence suggesting that ectopically produced PRL and/or locally released angiogenic peptides might play a causal role is provided.

FGF-2 stimulates migration of Kaposi's sarcoma-like vascular cells by HGF-dependent relocalization of the urokinase receptor.

The spindle-shaped cell line TTB was recently isolated from highly vascularized skin lesions of BKV/HIV-1 tat transgenic mice and shown to possess an autocrine loop for hepatocyte growth factor (HGF). We show that fibroblast growth factor-2 (FGF-2) stimulates TTB cell migration and promotes polarization of uPAR at the leading edge of migrating cells. FGF-stimulated TTB cells presented the typical migratory phenotype, with a triangular cell shape and concomitant breakdown of actin stress fibers and smooth muscle-specific actin isoform. FGF-2-stimulated migration was blocked by antibodies against urokinase-type plasminogen activator (uPA) or uPA receptor (uPAR) and by neutralizing anti-HGF antibodies. The latter also inhibited uPAR relocalization at the cell surface of FGF-2-treated TTB cells. This points to a crosstalk between FGF-2 and HGF that might mediate TTB cell migration by modulating the localization of cell surface uPAR.

Fibronectin modulates the effects of HIV-1 Tat on the growth of murine Kaposi's sarcoma-like cells through the down-regulation of tyrosine phosphorylation.

HIV-1 Tat plays a role in the pathogenesis of Kaposi's sarcoma. We therefore investigated the effect of Tat on the growth of murine Kaposi's sarcoma-like spindle (TTB) cells derived from dermal lesions. We observed that Tat and a peptide corresponding to the carboxyl-terminal region (Tat65-80) containing an RGD sequence inhibit TTB cell proliferation only when cells are cultured on fibronectin. This inhibitory effect correlates with redistribution of the alpha(v) integrin subunit on the surface of TTB cells and with down-regulation of tyrosine phosphorylation of specific substrates due to an increased tyrosine phosphatase activity. Indeed, phenylarsine oxide, a potent inhibitor of phosphotyrosine phosphatases, prevented the effects of Tat on TTB cells. We therefore argue that the action of Tat on TTB cells is mediated by the RGD motif through an integrin-based cell signaling pathway involving the activity of phosphotyrosine phosphatase(s), which would lead to a decrease in the levels of phosphotyrosine-containing proteins, among which is erk-2/p42MAPK.

Characterization of a saporin isoform with lower ribosome-inhibiting activity.

We have expressed in Escherichia coli five isoforms of saporin, a single-chain ribosome-inactivating protein (RIP). Translation inhibition activities of the purified recombinant polypeptides in vitro were compared with those of recombinant dianthin 30, a less potent and closely related RIP, and of ricin A chain. Dianthin 30, and a saporin isoform encoded by a cDNA from leaf tissue (SAP-C), both had about one order of magnitude lower activity in translation inhibition assays than all other isoforms of saporin tested. We recently demonstrated that saporin extracted from seeds of Saponaria officinalis binds to alpha2-macroglobulin receptor (alpha2MR; also termed low density lipoprotein-receptor-related-protein), indicating a general mechanism of interaction of plant RIPs with the alpha2MR system [Cavallaro, Nykjaer, Nielsen and Soria (1995) Eur. J. Biochem. 232, 165-171]. Here we report that SAP-C bound to alpha2MR equally well as native saporin. However, the same isoform had about ten times lower cytotoxicity than the other saporin isoforms towards different cell lines. This indicates that the lower cell-killing ability of the SAP-C isoform is presumably due to its altered interaction with the protein synthesis machinery of target cells. Since saporin binding to the alpha2MR is competed by heparin, we also tested in cell-killing experiments Chinese hamster ovary cell lines defective for expression of either heparan sulphates or proteoglycans. No differences were observed in cytotoxicity using native saporin or the recombinant isoforms. Therefore saporin binding to the cell surface should not be mediated by interaction with proteoglycans, as is the case for other alpha2MR ligands.

Spindle cells isolated from Kaposi's sarcoma-like lesions of BKV/tat-transgenic mice co-express markers of different cell types.

OBJECTIVE: To characterize murine spindle cells isolated from Kaposi's sarcoma-like skin lesions developed in BK virus (BKV)/tat-transgenic mice. METHODS: Kaposi's sarcoma-like spindle cells isolated from the lesions were propagated in vitro, and their phenotype was investigated using a panel of antibodies against various cell markers and angiogenic factors. Immunofluorescence and Western blot techniques were used. RESULTS: We observed co-expression of antigens specific for endothelial, smooth muscle and antigen-presenting cells, suggesting that cells from the TTB cell line represent poorly differentiated vascular precursors. Since TTB cells were derived from highly vascularized skin lesions, it is noteworthy that they synthesize a complex mixture of angiogenic factors, including fibroblast growth factor-2, vascular endothelial growth factor, placental growth factor, and hepatocyte growth factor. Due to their role in invasiveness and angiogenesis, we also observed the expression of urokinase plasminogen activator (uPA), uPA receptor, and plasminogen activator inhibitor-type 1 by TTB cells. CONCLUSIONS: Our results suggest that TTB cells share several features with human Kaposi's sarcoma spindle cells and can be a useful in vitro system to study the molecular mechanisms involved in Kaposi's sarcoma pathogenesis. Moreover, they synthesize a complex mixture of angiogenic factors and are growth-inhibited by the anti-angiogenic drug AGM-1470.

Targeting plant toxins to the urokinase and alpha 2-macroglobulin receptors.

We have conjugated the ribosome-inactivating protein (RIP) saporin to human urokinase-type plasminogen activator (uPA), and tested the uPA-saporin conjugate for cytotoxicity to uPA receptor-expressing cells. Unlike unconjugated uPA, saporin-conjugated uPA did not require any interaction with plasminogen activator inhibitors to be internalized. We have shown that saporin, as well as other RIPs, binds to the alpha 2-macroglobulin receptor (alpha 2 MR), a cell surface glycoprotein that endocytoses uPA-inhibitor complexes. Thus, the uPA receptor might present uPA saporm to alpha 2MR for internalization of the conjugate, a mechanism similar to that of uPA complexed to specific PA inhibitors. The binding of RIPs to alpha 2MR contrasts with previously proposed non-specific mechanisms for RIP entry into cells. The implications of the interactions between RIPs and alpha 2MR are discussed, with an emphasis on the role of alpha 2MR in mediating RIP immunogenicity and immune suppression.

Alpha 2-macroglobulin receptor mediates binding and cytotoxicity of plant ribosome-inactivating proteins.

It has been proposed that unconjugated type I ribosome-inactivating proteins (RIP) enter cells through passive mechanisms such as fluid-phase pinocytosis. However, some observations, such as the difference in sensitivity to type I RIP among different cell types, and the organ-specific toxicity of type I RIP, indicate a specific mechanism for the entry of these proteins into target cells. The alpha 2-macroglobulin receptor (alpha 2MR) is responsible for the binding and endocytosis of several ligands, including alpha 2-macroglobulin/proteinase complexes, plasminogen-activator-inhibitor complexes, apoE-enriched beta-very low density lipoproteins, and lipoprotein lipase. Here we demonstrate that saporin, a potent type I RIP, binds specifically to purified alpha 2MR and the binding is prevented by some alpha 2MR ligands. Moreover, the occupancy of specific ligand-binding sites on cell surface alpha 2MR decreases the cytotoxicity of saporin. The A chain of ricin, a type II RIP, also interacts with alpha 2MR. This, and the fact that saporin and ricin A chain both interact also with alpha 2-macroglobulin, indicates a general mechanism of complex interactions between RIP and cellular membranes that is mediated by alpha 2-macroglobulin and the alpha 2MR system.

A conjugate between human urokinase and saporin, a type-1 ribosome-inactivating protein, is selectively cytotoxic to urokinase receptor-expressing cells.

Urokinase-type plasminogen activator (uPA) confers invasive potential to transformed cells. Cancer cells express high numbers of uPA receptors (uPARs), which concentrate uPA activity at the invasive edge of cancer cells and the tumor mass. We synthesized a conjugate between human uPA and saporin (SAP), a ribosome-inactivating protein produced by Saponaria officinalis. Results of cell-killing assays showed that uPA is very effective at targeting saporin specifically to uPAR-expressing cells, whereas cell lines devoid of uPARs were not affected by the conjugate. Receptor-bound uPA is internalized only upon formation of a complex with one of its inhibitors (PAIs). However, our conjugate was highly cytotoxic even when the interaction between uPA and PAIs was prevented. Moreover, the alpha 2-macroglobulin receptor, which has been reported to mediate the internalization of uPA.PAI complexes, seems not to be involved in cell killing caused by the uPA.SAP conjugate. Thus, uPA.SAP might follow a mechanism of internalization different from that of unconjugated uPA complexed to PAIs, although still uPAR-mediated. Our results also suggest that alpha 2-macroglobulin and/or its receptor could mediate the internalization and cytotoxicity of unconjugated saporin, as it has been shown for other toxins.