A minority of carcinoma cells producing acidic fibroblast growth factor induces a community effect for tumor progression

scCross: a deep generative model for unifying single-cell multi-omics with seamless integration, cross-modal generation, and in silico exploration

Single-cell multi-omics data reveal complex cellular states, providing significant insights into cellular dynamics and disease. Yet, integration of multi-omics data presents challenges. Some modalities have not reached the robustness or clarity of established transcriptomics. Coupled with data scarcity for less established modalities and integration intricacies, these challenges limit our ability to maximize single-cell omics benefits. We introduce scCross, a tool leveraging variational autoencoders, generative adversarial networks, and the mutual nearest neighbors (MNN) technique for modality alignment. By enabling single-cell cross-modal data generation, multi-omics data simulation, and in silico cellular perturbations, scCross enhances the utility of single-cell multi-omics studies.

Triple-Negative Breast Cancer: Basic Biology and Immuno-Oncolytic Viruses

Triple-negative breast cancer (TNBC) is the most lethal subtype of breast cancer. TNBC diagnoses account for approximately one-fifth of all breast cancer cases globally. The lack of receptors for estrogen, progesterone, and human epidermal growth factor 2 (HER-2, CD340) results in a lack of available molecular-based therapeutics. This increases the difficulty of treatment and leaves more traditional as well as toxic therapies as the only available standards of care in many cases. Recurrence is an additional serious problem, contributing substantially to its higher mortality rate as compared to other breast cancers. Tumor heterogeneity also poses a large obstacle to treatment approaches. No driver of tumor development has been identified for TNBC, and large variations in mutational burden between tumors have been described previously. Here, we describe the biology of six different subtypes of TNBC, based on differential gene expression. Subtype differences can have a large impact on metastatic potential and resistance to treatment. Emerging antibody-based therapeutics, such as immune checkpoint inhibitors, have available targets for small subsets of TNBC patients, leading to partial responses and relatively low overall efficacy. Immuno-oncolytic viruses (OVs) have recently become significant in the pursuit of effective treatments for TNBC. OVs generally share the ability to ignore the heterogeneous nature of TNBC cells and allow infection throughout a treated tumor. Recent genetic engineering has allowed for the enhancement of efficacy against certain tumor types while avoiding the most common side effects in non-cancerous tissues. In this review, TNBC is described in order to address the challenges it presents to potential treatments. The OVs currently described preclinically and in various stages of clinical trials are also summarized, as are their strategies to enhance therapeutic potential.

The Unfolded Protein Response at the Tumor-Immune Interface

The tumor-immune interface has surged to primary relevance in an effort to understand the hurdles facing immune surveillance and cancer immunotherapy. Reports over the past decades have indicated a role for the unfolded protein response (UPR) in modulating not only tumor cell fitness and drug resistance, but also local immunity, with emphasis on the phenotype and altered function of immune cells such as myeloid cells and T cells. Emerging evidence also suggests that aneuploidy correlates with local immune dysregulation. Recently, we reported that the UPR serves as a link between aneuploidy and immune cell dysregulation in a cell nonautonomous way. These new findings add considerable complexity to the organization of the tumor microenvironment (TME) and the origin of its altered function. In this review, we summarize these data and also discuss the role of aneuploidy as a negative regulator of local immunity.

Cancer cell states and emergent properties of the dynamic tumor system

Phenotypic heterogeneity within malignant cells of a tumor is emerging as a key property of tumorigenesis. Recent work using single-cell transcriptomics has led to the identification of distinct cancer cell states across a range of cancer types, but their functional relevance and the advantage that they provide to the tumor as a system remain elusive. We present here a definition of cancer cell states in terms of coherently and differentially expressed gene modules and review the origins, dynamics, and impact of states on the tumor system as a whole. The spectrum of cell states taken on by a malignant population may depend on cellular lineage, epigenetic history, genetic mutations, or environmental cues, which has implications for the relative stability or plasticity of individual states. Finally, evidence has emerged that malignant cells in different states may cooperate or compete within a tumor niche, thereby providing an evolutionary advantage to the tumor through increased immune evasion, drug resistance, or invasiveness. Uncovering the mechanisms that govern the origin and dynamics of cancer cell states in tumorigenesis may shed light on how heterogeneity contributes to tumor fitness and highlight vulnerabilities that can be exploited for therapy.

Collapse of Intra-Tumor Cooperation Induced by Engineered Defector Cells

Anti-cancer therapies promote clonal selection of resistant cells that evade treatment. Effective therapy must be stable against the evolution of resistance. A potential strategy based on concepts from evolutionary game theory is to impair intra-tumor cooperation using genetically modified cells in which genes coding for essential growth factors have been knocked out. Such engineered cells would spread by clonal selection, driving the collapse of intra-tumor cooperation and a consequent reduction in tumor growth. Here, I test this idea in vitro in four cancer types (neuroendocrine pancreatic cancer, mesothelioma, lung adenocarcinoma and multiple myeloma). A reduction, or even complete eradication, of the producer clone and the consequent reduction in cell proliferation, is achieved in some but not all cases by introducing a small fraction of non-producer cells in the population. I show that the collapse of intra-tumor cooperation depends on the cost/benefit ratio of growth factor production. When stable cooperation among producer and non-producer cells occurs, its collapse can be induced by increasing the number of growth factors available to the cells. Considerations on nonlinear dynamics in the framework of evolutionary game theory explain this as the result of perturbation of the equilibrium of a system that resembles a public goods game, in which the production of growth factors is a cooperative phenotype. Inducing collapse of intra-tumor cooperation by engineering cancer cells will require the identification of growth factors that are essential for the tumor and that have a high cost of production for the cell.

Cooperative success in epithelial public goods games

Cancer cells obtain mutations which rely on the production of diffusible growth factors to confer a fitness benefit. These mutations can be considered cooperative, and studied as public goods games within the framework of evolutionary game theory. The population structure, benefit function and update rule all influence the evolutionary success of cooperators. We model the evolution of cooperation in epithelial cells using the Voronoi tessellation model. Unlike traditional evolutionary graph theory, this allows us to implement global updating, for which birth and death events are spatially decoupled. We compare, for a sigmoid benefit function, the conditions for cooperation to be favoured and/or beneficial for well-mixed and structured populations. We find that when population structure is combined with global updating, cooperation is more successful than if there were local updating or the population were well-mixed. Interestingly, the qualitative behaviour for the well-mixed population and the Voronoi tessellation model is remarkably similar, but the latter case requires significantly lower incentives to ensure cooperation.

IRE1α regulates macrophage polarization, PD-L1 expression, and tumor survival

In the tumor microenvironment, local immune dysregulation is driven in part by macrophages and dendritic cells that are polarized to a mixed proinflammatory/immune-suppressive phenotype. The unfolded protein response (UPR) is emerging as the possible origin of these events. Here we report that the inositol-requiring enzyme 1 (IRE1α) branch of the UPR is directly involved in the polarization of macrophages in vitro and in vivo, including the up-regulation of interleukin 6 (IL-6), IL-23, Arginase1, as well as surface expression of CD86 and programmed death ligand 1 (PD-L1). Macrophages in which the IRE1α/X-box binding protein 1 (Xbp1) axis is blocked pharmacologically or deleted genetically have significantly reduced polarization and CD86 and PD-L1 expression, which was induced independent of IFNγ signaling, suggesting a novel mechanism in PD-L1 regulation in macrophages. Mice with IRE1α- but not Xbp1-deficient macrophages showed greater survival than controls when implanted with B16.F10 melanoma cells. Remarkably, we found a significant association between the IRE1α gene signature and CD274 gene expression in tumor-infiltrating macrophages in humans. RNA sequencing (RNASeq) analysis showed that bone marrow-derived macrophages with IRE1α deletion lose the integrity of the gene connectivity characteristic of regulated IRE1α-dependent decay (RIDD) and the ability to activate CD274 gene expression. Thus, the IRE1α/Xbp1 axis drives the polarization of macrophages in the tumor microenvironment initiating a complex immune dysregulation leading to failure of local immune surveillance.

Evolution of cooperation in an epithelium

Cooperation is prevalent in nature, not only in the context of social interactions within the animal kingdom but also on the cellular level. In cancer, for example, tumour cells can cooperate by producing growth factors. The evolution of cooperation has traditionally been studied for well-mixed populations under the framework of evolutionary game theory, and more recently for structured populations using evolutionary graph theory (EGT). The population structures arising due to cellular arrangement in tissues, however, are dynamic and thus cannot be accurately represented by either of these frameworks. In this work, we compare the conditions for cooperative success in an epithelium modelled using EGT, to those in a mechanical model of an epithelium—the Voronoi tessellation (VT) model. Crucially, in this latter model, cells are able to move, and birth and death are not spatially coupled. We calculate fixation probabilities in the VT model through simulation and an approximate analytic technique and show that this leads to stronger promotion of cooperation in comparison with the EGT model.

Cooperation among cancer cells: applying game theory to cancer

Cell cooperation promotes many of the hallmarks of cancer via the secretion of diffusible factors that can affect cancer cells or stromal cells in the tumour microenvironment. This cooperation cannot be explained simply as the collective action of cells for the benefit of the tumour because non-cooperative subclones can constantly invade and free-ride on the diffusible factors produced by the cooperative cells. A full understanding of cooperation among the cells of a tumour requires methods and concepts from evolutionary game theory, which has been used successfully in other areas of biology to understand similar problems but has been underutilized in cancer research. Game theory can provide insights into the stability of cooperation among cells in a tumour and into the design of potentially evolution-proof therapies that disrupt this cooperation.

Insidious communication amongst cancer cells

The tumor microenvironment is home to various types of cognate and non-cognate cell interactions. Here we comment on a newly discovered form of intercellular communication, which is based on endoplasmic reticulum stress signaling. Through this mechanism transmitter cancer cells impart receiver cancer cells with resistance to secondary metabolic, pharmacologic and genotoxic stress, providing survival advantage. The implications of this finding are briefly discussed.

Evolutionary emergence of angiogenesis in avascular tumors using a spatial public goods game

Natural selection in cancer often results in the emergence of increasingly malignant tumor cells that display many if not all of the hallmarks of cancer. One of the most important traits acquired during cancer progression is angiogenesis. Tumor cells capable of secreting pro-angiogenic factors can be seen as cooperators where the improved oxygenation, nutrient delivery and waste disposal resulting from angiogenesis could be seen as a public good. Under this view, the relatively costly secretion of molecular signals required to orchestrate angiogenesis would be undertaken exclusively by cooperating tumor cells but the benefits of angiogenesis would be felt by neighboring tumor cells regardless of their contribution to the process. In this work we detail a mathematical model to better understand how clones capable of secreting pro-angiogenic factors can emerge in a tumor made of non-cooperative tumor cells. Given the importance of the spatial configuration of the tumor in determining the efficacy of the secretion of pro-angiogenic factors as well as the benefits of angiogenesis we have developed a spatial game theoretic approach where interactions and public good diffusion are described by graphs. The results show that structure of the population affects the evolutionary dynamics of the pro-angiogenic clone. Specifically, when the benefit of angiogenesis is represented by sigmoid function with regards to the number of pro-angiogenic clones then the probability of the coexistence of pro-angiogenic and angiogenesis-neutral clones increases. Our results demonstrate that pro-angiogenic clone equilibrates into clusters that appear from surrounding vascular tissues towards the center of tumor. These clusters appear notably less dense after anti-angiogenic therapy.

Do cell-autonomous and non-cell-autonomous effects drive the structure of tumor ecosystems?

By definition, a driver mutation confers a growth advantage to the cancer cell in which it occurs, while a passenger mutation does not: the former is usually considered as the engine of cancer progression, while the latter is not. Actually, the effects of a given mutation depend on the genetic background of the cell in which it appears, thus can differ in the subclones that form a tumor. In addition to cell-autonomous effects generated by the mutations, non-cell-autonomous effects shape the phenotype of a cancer cell. Here, we review the evidence that a network of biological interactions between subclones drives cancer cell adaptation and amplifies intra-tumor heterogeneity. Integrating the role of mutations in tumor ecosystems generates innovative strategies targeting the tumor ecosystem's weaknesses to improve cancer treatment.

The evolving paradigm of cell-nonautonomous UPR-based regulation of immunity by cancer cells

The endoplasmic reticulum (ER) stress response/unfolded protein response (UPR) has been thought to influence tumorigenesis mainly through cell-intrinsic, pro-survival effects. In recent years, however, new evidence has emerged showing that the UPR is also the source of cell-extrinsic effects, particularly directed at those immune cells within the tumor microenvironment. Here we will review and discuss this new body of information with focus on the role of cell-extrinsic effects on innate and adaptive immunity, suggesting that the transmission of ER stress from cancer cells to myeloid cells in particular is an expedient used by cancer cells to control the immune microenvironment, which acquires pro-inflammatory as well as immune-suppressive characteristics. These new findings can now be seen in the broader context of similar phenomena described in Caenorhabditis elegans, and an analogy with quorum sensing and 'community effects' in prokaryotes and eukaryotes can be drawn, arguing that a cell-nonautonomous UPR-based regulation of heterologous cells may be phylogenetically conserved. Finally, we will discuss the role of aneuploidy as an inducer of proteotoxic stress and potential initiator of cell-nonautonomous UPR-based regulation. In presenting these new views, we wish to bring attention to the cell-extrinsic regulation of tumor growth, including tumor UPR-based cell-nonautonomous signaling as a mechanism of maintaining tumor heterogeneity and resistance to therapy, and suggest therapeutically targeting such mechanisms within the tumor microenvironment.

Heterogeneity for IGF-II production maintained by public goods dynamics in neuroendocrine pancreatic cancer

The extensive intratumor heterogeneity revealed by sequencing cancer genomes is an essential determinant of tumor progression, diagnosis, and treatment. What maintains heterogeneity remains an open question because competition within a tumor leads to a strong selection for the fittest subclone. Cancer cells also cooperate by sharing molecules with paracrine effects, such as growth factors, and heterogeneity can be maintained if subclones depend on each other for survival. Without strict interdependence between subclones, however, nonproducer cells can free-ride on the growth factors produced by neighboring producer cells, a collective action problem known in game theory as the "tragedy of the commons," which has been observed in microbial cell populations. Here, we report that similar dynamics occur in cancer cell populations. Neuroendocrine pancreatic cancer (insulinoma) cells that do not produce insulin-like growth factor II (IGF-II) grow slowly in pure cultures but have a proliferation advantage in mixed cultures, where they can use the IGF-II provided by producer cells. We show that, as predicted by evolutionary game theory, producer cells do not go extinct because IGF-II acts as a nonlinear public good, creating negative frequency-dependent selection that leads to a stable coexistence of the two cell types. Intratumor cell heterogeneity can therefore be maintained even without strict interdependence between cell subclones. Reducing the amount of growth factors available within a tumor may lead to a reduction in growth followed by a new equilibrium, which may explain relapse in therapies that target growth factors.

Dynamics of growth factor production in monolayers of cancer cells and evolution of resistance to anticancer therapies

Tumor heterogeneity is well documented for many characters, including the production of growth factors, which improve tumor proliferation and promote resistance against apoptosis and against immune reaction. What maintains heterogeneity remains an open question that has implications for diagnosis and treatment. While it has been suggested that therapies targeting growth factors are robust against evolved resistance, current therapies against growth factors, like antiangiogenic drugs, are not effective in the long term, as resistant mutants can evolve and lead to relapse. We use evolutionary game theory to study the dynamics of the production of growth factors by monolayers of cancer cells and to understand the effect of therapies that target growth factors. The dynamics depend on the production cost of the growth factor, on its diffusion range and on the type of benefit it confers to the cells. Stable heterogeneity is a typical outcome of the dynamics, while a pure equilibrium of nonproducer cells is possible under certain conditions. Such pure equilibrium can be the goal of new anticancer therapies. We show that current therapies, instead, can be effective only if growth factors are almost completely eliminated and if the reduction is almost immediate.

IL-18 regulates melanoma VLA-4 integrin activation through a Hierarchized sequence of inflammatory factors

Very late antigen-4 (VLA-4) is frequently overexpressed on melanoma cells contributing to inflammation-dependent metastasis. Melanoma cell adhesion to endothelium via VLA-4-vascular cell adhesion molecule-1 (VCAM-1) interaction was used to study VLA-4 activation during melanoma cell response to inflammation. Cooperation among major inflammatory mediators was analyzed in melanoma cells exposed to single inflammatory factors in the presence of inhibitors for other assayed mediators. A stepwise cascade of hierarchized molecules heterogeneously made and used during melanoma response to IL-18, induced hydrogen peroxide (H2O2), in turn activating VLA-4 and melanoma cell adhesion to endothelium. The cascade involved prostaglandin E2 (PGE2) production from melanoma induced by IL-18-dependent tumor necrosis factor-α (TNFα); next, PGE2-induced IL-1β via vascular endothelial growth factor (VEGF) secretion, which in turn induced VLA-4 activation via cyclooxygenase 2-dependent H2O2. This sequence operated in IL-18R/VLA-4/VEGF-expressing murine (B16) and human (A375 and 883) melanomas, but not in those without this phenotype. Separation of active VLA-4-expressing B16 melanoma cells through immobilized VCAM-1 verified their higher IL-18R/TNFR1/VEGFR2 expression and metastatic growth than inactive VLA-4-expressing cells. However, cooperation among melanoma cell sub-populations with heterogeneous cytokine receptor levels may occur through VLA-4-stimulating factors, leading to intratumoral amplification of metastatic potential. Therefore, expression of the VLA-4-stimulating factor sequence may help to predict melanoma prometastatic risk, and offers therapeutic targets for metastatic melanoma deactivation through VLA-4 activation blockade.

Evolutionary game theory of growth factor production: implications for tumour heterogeneity and resistance to therapies

Background:

Tumour heterogeneity is documented for many characters, including the production of growth factors, one of the hallmarks of cancer. What maintains heterogeneity remains an open question that has implications for diagnosis and treatment, as drugs that target growth factors are susceptible to the evolution of resistance.

Methods:

I use evolutionary game theory to model collective interactions between cancer cells, to analyse the dynamics of the production of growth factors and the effect of therapies that reduce their amount.

Results:

Five types of dynamics are possible, including the coexistence of producer and non-producer cells, depending on the production cost of the growth factor, on its diffusion range and on the degree of synergy of the benefit it confers to the cells. Perturbations of the equilibrium mimicking therapies that target growth factors are effective in reducing the amount of growth factor in the long term only if the reduction is extremely efficient and immediate.

Conclusion:

Collective interactions within the tumour can maintain heterogeneity for the production of growth factors and explain why therapies like anti-angiogenic drugs and RNA interference that reduce the amount of available growth factors are effective in the short term but often lead to relapse. Alternative strategies for evolutionarily stable treatments are discussed.

The effect of vascular endothelial growth factor in the progression of bladder cancer and diabetic retinopathy

Bladder cancer and diabetic retinopathy is a major public health and economical burden worldwide. Despite its high prevalence, the molecular mechanisms that induce or develop bladder carcinomas and diabetic retinopathy progression are poorly understood but it might be due to the disturbance in balance between angiogenic factors such as VEGF and antiangiogenic factors such as pigment epithelium derived growth factor. VEGF is one of the important survival factors for endothelial cells in the process of normal physiological and abnormal angiogenesis and induce the expression of antiapoptotic proteins in the endothelial cells. It is also the major initiator of angiogenesis in cancer and diabetic retinopathy, where it is up-regulated by oncogenic expression and different type of growth factors. The alteration in VEGF and VEGF receptors gene and overexpression, determines a diseases phenotype and ultimately the patient's clinical outcome. However, expressional and molecular studies were made on VEGF to understand the exact mechanism of action in the genesis and progression of bladder carcinoma and diabetic retinopathy , but still how VEGF mechanism involve in such type of disease progression are not well defined. Some other factors also play a significant role in the process of activation of VEGF pathways. Therefore, further detailed analysis via molecular and therapeutic is needed to know the exact mechanisms of VEGF in the angiogenesis pathway. The detection of these types of diseases at an early stage, predict how it will behave and act in response to treatment through regulation of VEGF pathways. The present review aimed to summarize the mechanism of alteration of VEGF gene pathways, which play a vital role in the development and progression of bladder cancer and diabetic retinopathy.

Defining the role of cooperation in early tumor progression

Competition among cells has long been recognized as an important part of carcinogenesis. However, the role of cellular cooperation in cancer has been largely ignored. In this work, we investigated the role of cooperation in early tumor progression using a mathematical and agent-based modeling approach. We hoped to learn how the introduction of cooperative cells into a cell population would affect the dynamics of the system. We modeled the stem cell compartment of tissue using a spatial structure organized into cell patches, with stem cells able to replicate or leave the stem cell compartment through apoptosis or differentiation. The cells could also acquire mutations in three oncogenes and two tumor suppressor genes. Cooperative cells in our model provided a cooperative signal that increased the fitness of their immediate neighbors, but did not affect their own fitness. Running simulations of the model, we found that if cooperative cells are introduced into a cell population, they steadily proliferate and confer a growth advantage to the entire population. This leads us to conclude that providing a cooperative signal is likely to be under positive selective pressure. When cooperative ability and mutation are concurrently present in the same cells, the overall cell population experiences a significant growth advantage, much greater than with cooperation or mutation alone. This growth advantage is diminished if cells with only oncogene/tumor suppressor mutations are also present in the population, suggesting that the optimal scenario for tumor growth would be for cooperative cells to take over a cell population, and then for mutations in oncogenes and tumor suppressors to arise on a background of cooperation. We predict that cooperation is particularly important in the very early stages of carcinogenesis, when tissue is morphologically and histologically normal. Our results have implications for the screening and early diagnosis of cancer.

A novel single-chain variable fragment antibody against FGF-1 inhibits the growth of breast carcinoma cells by blocking the intracrine pathway of FGF-1

The fibroblast growth factors (FGFs) are important for embryo development, wound healing, hematopoiesis, and angiogenesis. FGF-1, a member of FGF family, is involved in both receptor-dependent pathways and an intracrine pathway. Studies have recently shown that FGF-1 is overexpressed in the early stages of several kinds of cancer. Thus, FGF-1 is a candidate for cancer immunotargeting. To study the potential use of therapeutic antibodies against FGF-1, a monoclonal hybridoma 1C9 secreting monoclonal antibody specific for FGF-1 was developed. Then, a single-chain variable fragment (scFv) antibody was genetically engineered from hybridama 1C9. The binding of the scFv1C9 to the antigen FGF-1 was demonstrated by ELISA and immunoprecipitation assays. Functional analysis showed that the overexpressed scFv1C9 in MCF-7 cells targeted endogenous FGF-1 and prevented the translocation of FGF-1 into the nucleus, resulting in the blockade of the intracrine pathway of FGF-1, which caused the G1 arrest by p21 up-regulation. These results suggest that the generated scFv1C9 is an effective inhibitor of the intracrine pathway of FGF-1 and has a potential application as anti-tumoral agent in breast cancer.

Dendritic cells transduced with TEM8 recombinant adenovirus prevents hepatocellular carcinoma angiogenesis and inhibits cells growth

Recent evidence suggested that angiogenesis played a pivotal role in the development of hepatocellular carcinoma cells (HCC), thus the therapy strategy targeting antiangiogenesis has been regarded as promising method for HCC therapy. Tumor endothelial marker 8 (TEM8) is a recently described protein that is preferentially expressed within tumor endothelium. However, the antiangiogenesis therapy of HCC based on TEM8 has not been reported. In this study, the recombinant adenovirus encoding TEM8 was constructed, and the DCs were transduced with the Ad-TEM8. In addition, the modified DCs were transferred into the BALB/c mice to determine whether DCs transduced with TEM8 could elicit a potent antitumor immunogenic response in vivo. The results demonstrated that DCs transduced with Ad-TEM8 induced specific CTLs effectively, which could secrete IFN-γ and lyse HCC. Furthermore, the modified DCs could effectively protect BALB/c mice from lethal challenges against HCC, reduce tumor growth and increase the mice life span by decreasing tumor vasculature density. These data suggest that the Ad-TEM8 modified DCs may induce antitumor immunity by disrupting tumor vasculature and may thus be used as an efficient therapy strategy to influence tumor development in clinical applications.

Recombinant adenovirus encoding vasohibin prevents tumor angiogenesis and inhibits tumor growth

Numerous lines of evidence have shown that angiogenesis plays a pivotal role in the development of tumors. Therefore anti-angiogenesis therapy represents a potentially promising approach to cancer therapy. Recently, a new inhibitor called vasohibin was discovered to operate as an intrinsic and highly specific feedback inhibitor in the process of angiogenesis. However, to date, reports on the antitumor and anti-angiogenic properties of vasohibin have been very limited. To explore the potential of vasohibin as an anti-angiogenesis therapeutic, we constructed a recombinant adenovirus encoding vasohibin. Our data showed that the recombinant adenovirus encoding vasohibin could prevent tumor angiogenesis and tumor growth. Notably, angiogenesis in the tumors was prevented without any apparent side-effects. Therefore, the findings suggested that the recombinant adenovirus encoding vasohibin might be valuable as a potential strategy for antitumor angiogenesis therapy in the clinic.

Trophoblast stem cell derivation, cross-species comparison and use of nuclear transfer: new tools to study trophoblast growth and differentiation

The trophoblast is a supportive tissue in mammals that plays key roles in embryonic patterning, foetal growth and nutrition. It shows an extensive growth up to the formation of the placenta. This growth is believed to be fed by trophoblast stem cells able to self-renew and to give rise to the differentiated derivatives present in the placenta. In this review, we summarize recent data on the molecular regulation of the trophoblast in vivo and in vitro. Most data have been obtained in the mouse, however, whenever relevant, we compare this model to other mammals. In ungulates, the growth of the trophoblast displays some striking features that make these species interesting alternative models for the study of trophoblast development. After the transfer of somatic nuclei into oocytes, studies in the mouse and the cow have both underlined that the trophoblast may be a direct target of reprogramming defects and that its growth seems specifically affected. We propose that the study of TS cells derived from nuclear transfer embryos may help to unravel some of the epigenetic abnormalities which occur therein.

Cancer as an evolutionary and ecological process

Neoplasms are microcosms of evolution. Within a neoplasm, a mosaic of mutant cells compete for space and resources, evade predation by the immune system and can even cooperate to disperse and colonize new organs. The evolution of neoplastic cells explains both why we get cancer and why it has been so difficult to cure. The tools of evolutionary biology and ecology are providing new insights into neoplastic progression and the clinical control of cancer.

Fibroblast growth factor signaling in tumorigenesis

Fibroblast growth factors and their signaling receptors have been associated with multiple biological activities, including proliferation, differentiation and motility. Consequently, they have evoked interest as candidate oncogenes with the potential to initiate and/or promote tumorigenesis. This has resulted in a large literature describing the presence of these growth factors and their receptors in cancer cell lines and primary tumors of diverse origin. However, it is only recently that compelling evidence has emerged to implicate the fibroblast growth factors (Fgfs) and their receptors in the genesis of human cancers. Here, we outline the model systems that demonstrate the potential oncogenic nature of Fgf signaling and summarise recent evidence that implicates aberrant Fgf signaling as important in the natural history of some common human cancers.

NBT-II carcinoma behaviour is not dependent on cell-cell communication through gap junctions

To study the mechanism(s) underlying the proliferation of heterogeneous cell populations within a solid tumour, the NBT-II rat bladder carcinoma system was used. It has been first investigated whether the different cell populations are coupled through gap junctions (GJIC). Cells overexpressing the Cx43 were generated to test for any tumour suppressive activity in vivo. To determine whether GJIC is essential for tumour proliferation and the establishment of a cooperative community effect, NBT-II cells that are incompetent for cell coupling were generated. The data report that (i) carcinoma cells expressing or not FGF-1 are coupled through GJIC in vitro and in coculture and express the gap junction protein Cx43, (ii) overexpression of Cx43 in these cells does not affect their in vitro coupling capacities and in vivo tumourigenic growth properties, (iii) inhibition of GJIC through antisense strategy has no in vivo obvious consequence on the tumour growth properties of the carcinoma, and (iv) the community effect between two carcinoma cell populations does not critically involve cell coupling through gap junctions.

Src kinase contributes to the metastatic spread of carcinoma cells

The involvement of Src kinase during carcinoma metastasis has been explored by using the NBT-II rat carcinoma cell line, which can be induced to scatter in vitro through Src activity. Here we show that Src activity was not required for growth of tumors derived from NBT-II cells injected into nude mice. In contrast, the presence of micrometastases was strictly dependent on Src, since the percentage of mice bearing metastases was dramatically reduced by the expression of a dominant-negative mutant of Src (SrcK-) or of Csk, the natural inhibitor of Src. Furthermore, metastatic cells originating from NBT-II cells displayed a Src activity higher than the parental cells, confirming that Src gives a selective advantage during the metastatic process. Finally, anatomopathological analysis of the primary tumors arising from NBT-II cells expressing Csk or SrcK- constructs revealed a highly differentiated epithelial phenotype contrasting with the poor differentiation of tumors derived from parental cells. The differentiated phenotype correlated with the presence of desmosomes at the cell periphery and the absence of vimentin intermediate filaments. Altogether, these data demonstrate that Src activity correlates with the loss of epithelial differentiation concomitantly with the increase of the metastatic potential of carcinoma cells.

FGF-1 and FGF-2 regulate the expression of E-cadherin and catenins in pancreatic adenocarcinoma

E-cadherin is a transmembrane protein that mediates Ca2+-dependent cell-cell adhesion and is implicated in a number of biologic processes, including cell growth and differentiation, cell recognition and cell sorting during development. We have previously demonstrated that both cell-cell adhesion and invasion are modulated by fibroblast growth factor (FGF)-1 and FGF-2 in a panel of pancreatic adenocarcinoma cell lines (BxPc3, T3M4 and HPAF). Here, we examine further the role of FGFs in the expression and activation of the E-cadherin/catenin system. We demonstrate that both FGF-1 and FGF-2 upregulate E-cadherin and beta-catenin at the protein level in the BxPc3 and HPAF cell lines and modestly in T3M4 cells. FGF-1 and FGF-2 facilitate the association of E-cadherin and alpha-catenin with the cytoskeleton, as demonstrated by the increase in the detergent-insoluble fraction of E-cadherin in BxPc3 and HPAF cells. Since the correct function of the E-cadherin/catenin complex requires its association with the cytoskeleton, our data suggest that FGF-1 and FGF-2 contribute to the integrity and thus the function of the complex. Furthermore, FGFs facilitate the assembly of the E-cadherin/catenin axis. The effect is associated with elevation of tyrosine phosphorylation of E-cadherin, alpha-catenin, beta-4051 mu-catenin and gamma-catenin, but not p120ctn. These findings indicate that the E-cadherin/catenin system is a target of the FGF/FGFR system and that coordinated signals from both systems may determine the ultimate biologic responses.

Oncogenes and angiogenesis: signaling three-dimensional tumor growth

Three-dimensional tumor growth is dependent on the perpetual recruitment of host blood vessels to the tumor site. This recruitment process (mainly via angiogenesis) is thought to be triggered, at least in part, by the very same set of genetic alterations (activated oncogenes, inactivated/lost tumor suppressor genes) as those responsible for other aspects of malignant transformation (e.g., aberrant mitogenesis, resistance to apoptosis). Potent oncogenes are able to deregulate expression of both angiogenesis stimulators and inhibitors in cancer cells. For example, mutant ras expression is associated with increased production of vascular endothelial growth factor (VEGF) and downregulation of thrombospondin-1 (TSP-1). Upregulation of VEGF and angiogenesis can also be induced by constitutive activation of other oncogenic proteins (e.g., EGFR, Raf, MEK, PI3K) acting at various levels on the Ras signaling pathway. The mode and the magnitude of such proangiogenic influences can be significantly modified by cell type (fibroblastic or epithelial origin), epigenetic factors (hypoxia, changes in cell density), and/or presence of additional genetic lesions (e.g., preceding loss of p16 or p53 tumor suppressor genes). Activated oncogenes (e.g., ras, src, HER-2) induce co-expression of angiogenic properties concomitantly with several highly selectable traits (increased mitogenesis, resistance to apoptosis), a circumstance that may accelerate selection of the angiogenic phenotype at the cell population level. On the other hand oncogene-induced reduction in growth requirements may also endow tumor cells with a diminished (albeit not abrogated) dependence on (close) proximity to blood vessels, i.e., with reduced vascular dependence. Thus, oncogenes can impact several interconnected aspects of cellular growth, survival, and angiogenesis. Experimental evidence suggests that, in principle, many of these properties (including angiogenesis) can be simultaneously suppressed (and tumor stasis or regression induced) by effective use of the specific oncogene antagonists and signal transduction inhibitors.

Co-inoculation of human and murine carcinoma cells induces reciprocal suppression of metastasis by both cell lines

The interactions of two cell lines having different metastatic properties, and the subsequent effects on dissemination were investigated using the chicken embryo metastasis assay. The highly aggressive human epidermoid cell line HEp-3 was tested alone or mixed with the mouse colon carcinoma cell line CL26 in this assay. When inoculated individually, each cell line forms experimental metastases in the chicken embryo, but only the HEp-3 cells give rise to spontaneous metastases. In embryos co-inoculated with both cell lines there was an overall reduction in metastatic burden in both the spontaneous and experimental metastasis assays. Furthers studies revealed that CL26 cells, when co-inoculated with HEp-3 cells did not acquire the ability to spontaneously metastasize. However, in the presence of CL26 cells, spontaneous HEp-3 metastasis was reduced. Intravenous co-inoculation of HEp-3 and CL26 cells also resulted in a reciprocal suppression of experimental metastasis by both cell lines. These studies demonstrate that the interactions of adjacent, phenotypically different tumor cells can have a suppressive effect on dissemination of one or both cell types.

aFGF immunoreactivity in prostate cancer and its co-localization with bFGF and FGF8

Fibroblast growth factors (FGFs) have been implicated in the development of numerous malignancies including prostate cancer. In a pilot study it has been shown that FGF8 mRNA is up-regulated in prostate cancer. The aim of the present study was to determine whether aFGF and bFGF were co-expressed with FGF8 in human prostate cancer. Twenty-nine cases of prostate cancer of different histological grades were examined. Immunohistochemical analysis was employed to study aFGF and bFGF expression. In the light of the results, aFGF immunoreactivity was studied in a further 43 cases. aFGF and bFGF immunoreactivity was identified in the cytoplasm of the malignant prostatic epithelium. aFGF was overexpressed in 62/72 (86.1 per cent) cases and bFGF in 19/29 (65.5 per cent). High levels of aFGF immunoreactivity were noted in areas of high-grade prostatic intraepithelial neoplasia (PIN). In this series, aFGF immunoreactivity was most commonly observed and correlated closely with Gleason score and tumour stage ( p=0.007 and 0.007, respectively). Co-localization of aFGF, bFGF, and FGF8 was detected in 9/29 (31.0 per cent) cases. There was a significant correlation between aFGF and FGF8 expression. In conclusion, aFGF, bFGF, and FGF8 are co-localized in human prostate cancer; they may have a synergistic effect in prostate cancer growth and progression.

Nuclear 24 kD fibroblast growth factor (FGF)-2 confers metastatic properties on rat bladder carcinoma cells

The tumorigenic and metastatic properties of rat bladder carcinoma NBT-II cells transfected with a cDNA encoding the 24 kD nuclear isoform of human fibroblast growth factor-2 (FGF-2) were analysed and compared with those cells producing the 18 kD cytoplasmic isoform FGF-2. In transfected clones, 24 kD FGF-2 was found in the nucleus, and no FGF-2 was secreted. RT-PCR analysis showed no upregulation of FGF-2-specific receptor FGFR2c expression in these proliferating transfected cells. A shorter latency period for in vivo tumor formation and abundant spontaneous lung metastases were only seen if nuclear FGF-2-producing cells were injected subcutaneously into nude mice. Intravenous injection of 24 kD FGF-2-producing cells led to extensive experimental lung metastases whereas injection of control NBT-II cells or 18 kD FGF-2-producing cells did not. As FGF-2-producing cells have no specific FGF-2 receptors, our results suggest that the 24 kD FGF-2 has nuclear targets, and activates metastatic property of carcinoma cells via a mechanism other than the conventional FGF receptor-mediated signaling pathway.

Collateral expression of proangiogenic and tumorigenic properties in intestinal epithelial cell variants selected for resistance to anoikis

Although in vitro anchorage-independent growth is widely used as a marker of cell transformation, the biological implications of this trait are poorly understood. We previously demonstrated that enforced anchorage-independent growth of a nontumorigenic, immortalized epithelial cell line (IEC-18) in multicellular spheroid culture results in massive apoptotic cell death. This death process, termed anoikis, is prevented by expression of transforming oncogenes, which also confer tumorigenic competence. This study examines whether acquisition of an anoikis-resistant phenotype is causally related to the tumorigenic capacity of transformed epithelial cells. Parental IEC-18 cells were subjected to 10 cycles of selection for survival in speroid culture. Unlike parental cells, the resulting anoikis-resistant variants (AR1.10 and AR2.10) formed relatively large tumors in nude mice. Both anoikis-resistant sublines displayed upregulated expression of vascular endothelial growth factor (VEGF), a potent angiogenesis stimulator. VEGF121 overexpression alone did not induce tumorigenic conversion of parental IEC-18 cells, which remained highly susceptible to anoikis. We postulate that both anoikis-resistance and angiogenic-competence contribute to tumor formation. Development of anoikis-resistance can be then viewed as a precondition for expression of the tumorigenic phenotype. Our results suggest that even when angiogenesis is not a rate limiting factor (e.g. in vitro) the selective pressures of solid tumor-like, 3-dimensional growth conditions favoring anoikis resistance result in collateral induction of a proangiogenic phenotype.

The community effect in FGF-1 mediated tumor progression of a rat bladder carcinoma does not involve a direct paracrine signaling

A community effect was found to occur between heterogeneous tumor cell populations leading to an overall increased tumorigenicity without a clonal dominance of the more tumorigenic clone. In the rat bladder carcinoma cell line NBT-II, this effect appears mediated by the Fibroblast Growth Factor-1 (FGF-1) through either a direct or an indirect signaling pathway. Neovascularization induced by FGF-1 was found not to be responsible for the community effect. The present study shows that the community effect does not involve a direct FGF-1 signaling since tumor cells expressing a dominant-negative FGF receptor mutant were still responding to the highly tumorigenic FGF-1 expressing cells. Tumors arising from inoculates of the FGF-1 producing NBT-II cells mixed with non tumorigenic epithelial MDCK cells contain only the tumorigenic cells indicating that MDCK cells may exerce a helper effect for the growth of the tumor not dependant on their own growth. Therefore the helper function of MDCK cells must be distinguished from a community effect where the contribution of low tumorigenic cells not only provides an in vivo growth advantage to few highly tumorigenic cells but become themselves highly tumorigenic indicating that the community effect may require cell-cell specific cooperativity independent from an helper effect.

Fibroblast growth factors and their receptors

Fibroblast growth factors (FGFs) represent a group of polypeptide mitogens eliciting a wide variety of responses depending upon the target cell type. The knowledge of the cell surface receptors mediating the effects of FGFs has recently expanded remarkably. The complexity of the FGF family and the FGF-induced responses is reflected in the diversity and redundancy of the FGF receptors. In this review, a number of biochemical characteristics and biological properties of the FGF family and its receptors are described and their expression both in normal tissues and in tumours is discussed. Finally we speculate on the targetting of growth inhibition agents to tumours through FGF receptors.

The cellular basis of tumor progression

Variability in disease presentation and course is a hallmark of cancer. Variability is seen among similarly diagnosed cancers in different patients or animal hosts and in the same cancer at different periods of time. This latter type of variability, termed "tumor progression," was defined by Foulds in a series of six rules that describe the independent behavior of individual cancers and the independent evolution of different cancer characteristics. Tumor progression is believed to result from variability among subpopulations of tumor cells within individual cancers and from selection of these subpopulations by conditions within the cancer environment, such that different subpopulations come to prominence over the course of cancer development and growth. Interactions among subpopulations, however, modulate tumor behavior as well as tumor evolution. The leading hypothesis for the origin of tumor subpopulations is the genetic instability of cancer cells. There are a number of possible mechanisms of genetic instability, some internal to cancer cells (mutation, amplification, mutator phenotypes, DNA repair deficiencies) and some present in the tumor microenvironment (endogenous mutagens). There are also potential epigenetic mechanisms of variability, including alterations in gene regulation, differentiation, adaptation, and cell fusion. Regardless of mechanism, the heterogeneity of tumor subpopulations poses a number of challenges to the practice of cancer research, including the design of reproducible and meaningful experiments. Tumor heterogeneity also has significant consequences for the clinical assessment of tumor prognosis and the development of effective treatment regimens.

Adenovirus-mediated gene transfer of fibroblast growth factor-1: angiogenesis and tumorigenicity in nude mice

Gene transfer of angiogenic growth factors with replication-deficient recombinant adenovirus (Ad) vectors may provide a new approach to the treatment of ischemic diseases. To determine if Ad-infected cells could stimulate angiogenesis in vivo and to assess the tumorigenicity of cells infected with these vectors, NIH3T3 fibroblasts infected with Ad vectors coding for human acidic fibroblast growth factor (aFGF-1) were used in angiogenic and tumorigenic assays. Infected cells induced a strong angiogenic response in vivo, while cells infected with control virus did not. Stable 3T3 transfectants expressing the FGF-1 gene were also highly angiogenic and exhibited growth in soft agar, while Ad-infected cells did not. Ad-infected cells grew transiently in nude mice, whereas 3T3 transfectants formed large tumors which grew exponentially. Extrapolation of cell dose-response curves showed that a minimum of 1.5 x 10(4) infected cells were required for transient tumor cell growth in vivo. Ad-infected cells cultured in vitro for 30 days lost their invasive phenotype and the ability for transient cell growth in nude mice. Thus, phenotypic changes induced by Ad-mediated gene transfer of FGF-1 are transient both in vitro and in vivo, suggesting that these Ad vectors do not have tumorigenic potential. Stimulation of angiogenesis by Ad-infected cells may be useful for the evaluation of anti-angiogenic and anti-tumor agents.

The community effect and ectoderm-mesoderm interaction in Xenopus muscle differentiation

Community effects are believed to play an important role in the patterning of many tissues during development. They involve an interaction between neighbouring equivalent cells that is necessary for them to proceed to their fully differentiated state. However, the mechanisms underlying these effects remain unclear. In this paper, diffusion-based mathematical models are constructed and analysed in order to study possible mechanisms for the community effect in Xenopus muscle differentiation. These models differ from each other in the assumptions that are made about the nature of an inhibitory effect that ectodermal tissue has been observed to have on muscle differentiation. It is possible to construct consistent models based on all the forms of inhibition considered. However, each model requires the diffusible factors on which it is based to have different properties. The current data from tissues reaggregate experiments are insufficient to determine the mechanisms underlying the community effect; the work presented here suggests that quantitative analysis of a further series of reaggregate experiments will make it possible to distinguish between the proposed mechanisms.

Pleiotropic over-expression of multidrug-resistance-related genes is correlated to MYCN and max mRNA accumulation during tumour progression in the IGR-N-91 human neuroblastoma model

An experimental model of advanced human neuroblastoma, IGR-N-91, which is able to disseminate in the nude mouse, has been described. The present study was designed to ascertain which cell population from the IGR-N-91 primary tumour actually disseminates throughout the animals. In s.c. IGR-N-91 tumour xenografts, 3 areas, called pearly, vascularized and haemorrhagic, depending on the presence of blood vessels and haemorrhagic suffusions, were consistently observed and independently resected. Molecular analysis of tumour materials revealed a significant increase in MYCN and max gene transcript levels in the haemorrhagic area, as compared with the pearly and vascularized areas. Given the growth kinetics observed both in vitro and in vivo, and the DNA flow-cytometry profiles of tumour cells obtained from the haemorrhagic area, this transcriptional increase did not appear to be associated with enhanced proliferation. In this area of the tumours, multidrug-resistance-related genes, i.e., MDRI, MRP, GST-pi and topoisomerase II alpha were activated concomitantly with MYCN and max genes. The same observations were made, except for the topoisomerase-II alpha gene, when sub-lines derived from metastases were compared with that derived from the primary tumour. These data demonstrate that over-expression of several genes determining the multi-drug-resistance phenotype precedes the metastatic spread of IGR-N-91 NB tumour cells in the nude mouse. Data also suggest that the cell sub-population exhibiting this pleiotropic over-expression within the primary tumour undergoes selection during metastatic dissemination.

Fibroblast growth factor 1 and fibroblast growth factor 2 immunoreactivity in gastrointestinal tumours

Acidic and basic fibroblast growth factors (FGF-1 and FGF-2) are mitogenic polypeptides that may play a role in autocrine and paracrine growth control of malignant tumours. We have examined the expression of FGF-1 and FGF-2 in a series of 41 colorectal tumours (24 adenomas, 17 adenocarcinomas) and 50 gastric adenocarcinomas (23 intestinal, 27 diffuse), using immunohistochemistry. Whereas the FGF-1 distribution was cytoplasmic, FGF-2 was restricted to the nuclei of the epithelial cells. FGF-1 immunoreactivity was detected in all samples (100 per cent), whereas FGF-2 immunoreactivity was seen in 17 adenomas (71 per cent), 13 colorectal carcinomas (76 per cent), and 29 gastric carcinomas (58 per cent). Compared with the normal mucosa, FGF-1 was overexpressed in 42 per cent of colorectal adenomas, 76 per cent of colorectal cancers, and 54 per cent of gastric cancers. Conversely, FGF-2 expression was reduced in 16 (66 per cent), 8 (47 per cent), and 40 (80 per cent) adenomas and colorectal and gastric samples, respectively. We found a significant correlation only between reduced FGF-2 and gastric tumour grade. These data indicate that FGF-1 overexpression occurs in a large proportion of human colorectal and gastric cancers. This may play a role in the progression of these tumours. The topographic variation in FGF-2 expression between normal (nuclear) and tumour (cytoplasmic) cells implies a corresponding functional change that may in turn facilitate tumour growth.

Five newly established oesophageal carcinoma cell lines: phenotypic and immunological characterization

The derivation of permanent cell lines from 40 resected oesophageal carcinomas has been attempted. Five long-term lines have been established from three adenocarcinomas, one mixed carcinoma and one squamous carcinoma. Molecular and cellular analyses have been carried out on the lines and clones derived from them. Karyotype analysis indicates genetic variation among the clones. HLA-A, -B and -C is expressed constitutively, but not HLA-DR. ICAM-1-expressing phenotypes may have arisen during adaptation to long-term culture. All lines are capable of response to interferon-gamma (IFN-gamma) and all produce transforming growth factor beta 1 (TGF-beta 1). Two lines are resistant to the inhibitory growth effects of the latter, possibly contributing to malignancy. It is anticipated that these lines, originating from histologically different carcinomas, will provide a valuable, continuous resource for the investigation and treatment of these aggressive tumours.

Molecular and cellular basis of cancer invasion and metastasis: implications for treatment

In the past decade significant advances in establishing the underlying biological mechanisms of tumour invasion and metastasis have been made. Some of the triggering factors and genes relevant to metastatic spread have been identified. Advances have also been made in understanding the signal transduction pathways involved in invasion and metastasis. This increased comprehension of the malignant metastatic process has enabled new antimetastatic strategies to be devised. This review summarizes progress in these areas and discusses the implications for the treatment of metastasis.

Selective interactions between mammary epithelial cells and fibroblasts in co-culture

Paracrine interactions between breast-cancer cells (MCF7) and stromal fibroblasts were studied in relation to the presence of steroid hormones, using co-cultures in which the 2 populations were separated by a microporous membrane. Densities and DNA-synthesis rates of the co-existing populations were interrelated. Proliferation was, therefore, viewed as the cumulative result of several factors, some of which are non-specific, e.g., are density-dependent, and some are specifically related to the feeders' origin and/or to culture conditions. Specific effects were measured and evaluated by stepwise analysis of covariance. MCF7 stimulated proliferation of fibroblasts differentially. Malignant-tumour fibroblasts were stimulated more than non-pathological ones. The magnitude of these effects was dependent on the presence of steroids. A similar analytical method was used for evaluating differential stromal influences on 4 epithelial phenotypic characters commonly used as prognostic markers. The estrogen-receptor, progesterone-receptor, pS2 and cathepsine-D phenotypes of MCF7, as well as their interrelations, were dependent on the origin of the fibroblasts, i.e., embryonic or adult, normal or tumoral.

A scatter factor-like factor is produced by a metastatic variant of a rat bladder carcinoma cell line

The rat bladder carcinoma epithelial NBT-II cell line undergoes, in vitro, a morphological transition to a fibroblast-like state in the presence of different growth factors. We have selected, in vivo, a metastatic clone, designated M-NBT-II, which has a mesenchymal phenotype and secretes into the culture medium a factor able to dissociate epithelial clusters of NBT-II or MDCK cells. This factor was designated scatter factor-like (SFL) by analogy to the HGF/SF, which has the same dissociating effect in these two cell lines. Here, we show that SFL factor and HGF/SF are different factors: (i) no HGF/SF transcripts could be detected using either specific rat HGF/SF cDNA probes or PCR; (ii) blocking antibodies against rat HGF/SF do not inhibit the SFL activity; and (iii) crude culture medium or partially purified SFL factor-containing fractions do not induce MDCK tubulogenesis, a biological assay that is specific for HGF/SF activity in vitro. We report the partial purification of the SFL factor, based on ion exchange and reverse-phase chromatography. The results indicate that the M-NBT-II metastatic variant secretes a dissociating factor sharing some common biological properties with the HGF/SF, which suggests that the SFL factor is a member of the HGF/SF family and may be involved in tumor progression.