Reversible lysine-specific demethylase 1 antagonist HCI-2509 inhibits growth and decreases c-MYC in castration- and docetaxel-resistant prostate cancer cells.

BACKGROUND: Lysine-specific demethylase 1 (LSD1 or KDM1A) overexpression correlates with poor survival and castration resistance in prostate cancer. LSD1 is a coregulator of ligand-independent androgen receptor signaling promoting c-MYC expression. We examined the antitumor efficacy of LSD1 inhibition with HCI-2509 in advanced stages of prostate cancer. METHODS: Cell survival, colony formation, histone methylation, c-MYC level, c-MYC expression, cell cycle changes and in vivo efficacy were studied in castration-resistant prostate cancer cells upon treatment with HCI-2509. In vitro combination studies, using HCI-2509 and docetaxel, were performed to assess the synergy. Cell survival, colony formation, histone methylation and c-myc levels were studied in docetaxel-resistant prostate cancer cells treated with HCI-2509. RESULTS: HCI-2509 is cytotoxic and inhibits colony formation in castration-resistant prostate cancer cells. HCI-2509 treatment causes a dose-dependent increase in H3K9me2 (histone H3lysine 9) levels, a decrease in c-MYC protein, inhibition of c-MYC expression and accumulation in the G0/G1 phase of the cell cycle in these cells. PC3 xenografts in mice have a significant reduction in tumor burden upon treatment with HCI-2509 with no associated myelotoxicity or weight loss. More synergy is noted at sub-IC50 (half-maximal inhibitory concentration) doses of docetaxel and HCI-2509 in PC3 cells than in DU145 cells. HCI-2509 has growth-inhibitory efficacy and decreases the c-myc level in docetaxel-resistant prostate cancer cells. CONCLUSIONS: LSD1 inhibition with HCI-2509 decreases the c-MYC level in poorly differentiated prostate cancer cell lines and has a therapeutic potential in castration- and docetaxel-resistant prostate cancer.

A phase I/II study of decitabine in combination with panitumumab in patients with wild-type (wt) KRAS metastatic colorectal cancer.

PURPOSE: KRAS mutations are predictive of lack of response to monoclonal antibodies (mAb) against EGFR in metastatic colorectal cancer (mCRC). Most wild-type KRAS patients, however, are also resistant. Retrospective data suggest that EGFR silencing play a role in resistance to therapy. We conducted a study to evaluate the safety and efficacy of decitabine (a hypomethylating agent) in combination with panitumumab (mAb against EGFR) in mCRC patients. EXPERIMENTAL DESIGN: 20 patients with wild-type KRAS mCRC were included in this phase I/II study. Patients were treated with decitabine at 45 mg/m(2) IV over 2 h on day 1 and 15 and panitumumab 6 mg/kg IV over 1 h on day 8 and 22 every 28 days. Blood samples were collected at baseline, day 8, 15 and 22. Quantitative polymerase chain reaction was used to measure promoter-specific methylation in peripheral-blood cells (PBMCs). RESULTS: The most common adverse events were grade 1-2 (rash and hypomagnesemia); 3 (16 %) patients had grade III-IV neutropenia including one patient with neutropenic fever. Two of 20 patients (10 %) had a partial response. Both had previously received cetuximab. Ten patients had stable disease (3 of them longer than 16 weeks). Decreased methylation of the MAGE promoter was not evidenced in PBMCs. CONCLUSIONS: The combination of decitabine and panitumumab was well tolerated and showed activity in wild-type KRAS mCRC patients previously treated with cetuximab. Target modulation in surrogate tissues was not achieved and tumor biopsies were not available. Future studies evaluating hypomethylating agents in combination with EGFR mAb in patients with mCRC are warranted.

A genomic approach to predict synergistic combinations for breast cancer treatment.

We leverage genomic and biochemical data to identify synergistic drug regimens for breast cancer. In order to study the mechanism of the histone deacetylase (HDAC) inhibitors valproic acid (VPA) and suberoylanilide hydroxamic acid (SAHA) in breast cancer, we generated and validated genomic profiles of drug response using a series of breast cancer cell lines sensitive to each drug. These genomic profiles were then used to model drug response in human breast tumors and show significant correlation between VPA and SAHA response profiles in multiple breast tumor data sets, highlighting their similar mechanism of action. The genes deregulated by VPA and SAHA converge on the cell cycle pathway (Bayes factor 5.21 and 5.94, respectively; P-value 10(-8.6) and 10(-9), respectively). In particular, VPA and SAHA upregulate key cyclin-dependent kinase (CDK) inhibitors. In two independent datasets, cancer cells treated with CDK inhibitors have similar gene expression profile changes to the cellular response to HDAC inhibitors. Together, these results led us to hypothesize that VPA and SAHA may interact synergistically with CDK inhibitors such as PD-033299. Experiments show that HDAC and CDK inhibitors have statistically significant synergy in both breast cancer cell lines and primary 3-dimensional cultures of cells from pleural effusions of patients. Therefore, synergistic relationships between HDAC and CDK inhibitors may provide an effective combinatorial regimen for breast cancer. Importantly, these studies provide an example of how genomic analysis of drug-response profiles can be used to design rational drug combinations for cancer treatment.

The alternative translation of synaptotagmin 1 mediates the non-classical release of FGF1.

Although the extravesicular p40 domain of the transmembrane protein, p65 synaptotagmin (Syt) 1, is essential for the non-classical export of the signal peptide-less structure, FGF1, it was not possible to identify a specific intracellular protease responsible for the processing of p65 Syt1. Surprisingly, analysis of the p65 Syt1 coding sequence revealed the presence of two potential alternative ATG codons corresponding to Met103 and Met113 both of which were flanked by Kozak sequences. Indeed, in vitro translation of a Met103Ile but not a Met113Ile p65 Syt1 point mutant exhibited reduced expression of p40 Syt1 and the double p65 Syt1 Met103Ile and Met113Ile point mutant was unable to translate the p40 Syt1 isoform. Since the expression of the p65 Syt1 double point mutant inhibited the stress-induced release of FGF1, it is likely that the alternative translation of the p65 Syt1 transcript at Met103 may be involved in the generation of intracellular p40 Syt1, a critical component of the FGF1 release pathway.

Copper induces the assembly of a multiprotein aggregate implicated in the release of fibroblast growth factor 1 in response to stress.

Fibroblast growth factor (FGF) 1 is known to be released in response to stress conditions as a component of a multiprotein aggregate containing the p40 extravescicular domain of p65 synaptotagmin (Syt) 1 and S100A13. Since FGF1 is a Cu2+-binding protein and Cu2+ is known to induce its dimerization, we evaluated the capacity of recombinant FGF1, p40 Syt1, and S100A13 to interact in a cell-free system and the role of Cu2+ in this interaction. We report that FGF1, p40 Syt1, and S100A13 are able to bind Cu2+ with similar affinity and to interact in the presence of Cu2+ to form a multiprotein aggregate which is resistant to low concentrations of SDS and sensitive to reducing conditions and ultracentrifugation. The formation of this aggregate in the presence of Cu2+ is dependent on the presence of S100A13 and is mediated by cysteine-independent interactions between S100A13 and either FGF1 or p40 Syt1. Interestingly, S100A13 is also able to interact in the presence of Cu2+ with Cys-free FGF1 and this observation may account for the ability of S100A13 to export Cys-free FGF1 in response to stress. Lastly, tetrathiomolybdate, a Cu2+ chelator, significantly represses in a dose-dependent manner the heat shock-induced release of FGF1 and S100A13. These data suggest that S100A13 may be involved in the assembly of the multiprotein aggregate required for the release of FGF1 and that Cu2+ oxidation may be an essential post-translational intracellular modifier of this process.

S100A13 participates in the release of fibroblast growth factor 1 in response to heat shock in vitro.

S100A13, a member of the S100 gene family of Ca(2+)-binding proteins has been previously characterized as a component of a brain-derived heparin-binding multiprotein aggregate/complex containing fibroblast growth factor 1 (FGF1). We report that while expression of S100A13 in NIH 3T3 cells results in the constitutive release of S100A13 into the extracellular compartment at 37 degrees C, co-expression of S100A13 with FGF1 represses the constitutive release of S100A13 and enables NIH 3T3 cells to release S100A13 in response to temperature stress. S100A13 release in response to stress occurs with kinetics similar to that observed for the stress-induced release of FGF1, but S100A13 expression is able to reverse the sensitivity of FGF1 release to inhibitors of transcription and translation. The release of FGF1 and S100A13 in response to heat shock results in the solubility of FGF1 at 100% (w/v) ammonium sulfate saturation, and the expression of a S100A13 deletion mutant lacking its novel basic residue-rich domain acts as a dominant negative effector of FGF1 release in vitro. Surprisingly, the expression of S100A13 also results in the stress-induced release of a Cys-free FGF1 mutant, which is normally not released from NIH 3T3 cells in response to heat shock. These data suggest that S100A13 may be a component of the pathway for the release of the signal peptide-less polypeptide, FGF1, and may involve a role for S100A13 in the formation of a noncovalent FGF1 homodimer.

Amlexanox reversibly inhibits cell migration and proliferation and induces the Src-dependent disassembly of actin stress fibers in vitro.

Amlexanox binds S100A13 and inhibits the release of fibroblast growth factor 1 (FGF1). Because members of the S100 gene family are known to be involved with the function of the cytoskeleton, we examined the ability of amlexanox to modify the cytoskeleton and report that amlexanox induces a dramatic reduction in the presence of actin stress fibers and the appearance of a random, non-oriented distribution of focal adhesion sites. Correspondingly, amlexanox induces the complete and reversible non-apoptotic inhibition of cell migration and proliferation, and although amlexanox does not induce either the down-regulation of F-actin levels or the depolymerization of actin filaments, it does induce the tyrosine phosphorylation of cortactin, a Src substrate known to regulate actin bundling. In addition, a dominant negative form of Src is able to partially rescue cells from the effect of amlexanox on both the actin cytoskeleton and cell migration. In contrast, the inhibition of cell proliferation by amlexanox correlates with the inhibition of cyclin D1 expression without interference of the receptor tyrosine kinase/mitogen-activated protein kinase signaling pathway. Last, the ability of amlexanox to inhibit FGF1 release is reversible and correlates with the restoration of the actin cytoskeleton, suggesting a role for the actin cytoskeleton in the FGF1 release pathway.

Identification of specific molecular structures of human immunodeficiency virus type 1 Tat relevant for its biological effects on vascular endothelial cells.

Human immunodeficiency virus type 1 (HIV-1) Tat transactivates viral genes and is released by infected cells, acting as a soluble mediator. In endothelial cells (EC), it activates a proangiogenic program by activating vascular endothelial growth factor receptor type 2 (VEGFR-2) and integrins. A structure-activity relationship study was performed by functional analysis of Tat substitution and deletion variants to define the Tat determinants necessary for EC activation. Variants were made (i) in the basic and (ii) in the cysteine-rich domains and (iii) in the C-terminal region containing the RGD sequence required for integrin recognition. Our results led to the following conclusions. (i) Besides a high-affinity binding site corresponding to VEGFR-2, EC express low-affinity binding sites. (ii) The basic and the cysteine-rich variants bind only to the low-affinity binding sites and do not promote tyrosine phosphorylation of VEGFR-2. Furthermore, they have a reduced ability to activate EC in vitro, and they lack angiogenic activity. (iii) Mutants with mutations in the C-terminal region are partially defective for in vitro biological activities and in vivo angiogenesis, but they activate VEGFR-2 as Tat wild type. In conclusion, regions encoded by the first exon of tat are necessary and sufficient for activation of VEGFR-2. However, the C-terminal region, most probably through RGD-mediated integrin engagement, is indispensable for full activation of an in vitro and in vivo angiogenic program.

Role of alphavbeta3 integrin in the activation of vascular endothelial growth factor receptor-2.

Interaction between integrin alphavbeta3 and extracellular matrix is crucial for endothelial cells sprouting from capillaries and for angiogenesis. Furthermore, integrin-mediated outside-in signals co-operate with growth factor receptors to promote cell proliferation and motility. To determine a potential regulation of angiogenic inducer receptors by the integrin system, we investigated the interaction between alphavbeta3 integrin and tyrosine kinase vascular endothelial growth factor receptor-2 (VEGFR-2) in human endothelial cells. We report that tyrosine-phosphorylated VEGFR-2 co-immunoprecipitated with beta3 integrin subunit, but not with beta1 or beta5, from cells stimulated with VEGF-A165. VEGFR-2 phosphorylation and mitogenicity induced by VEGF-A165 were enhanced in cells plated on the alphavbeta3 ligand, vitronectin, compared with cells plated on the alpha5beta1 ligand, fibronectin or the alpha2beta1 ligand, collagen. BV4 anti-beta3 integrin mAb, which does not interfere with endothelial cell adhesion to vitronectin, reduced (i) the tyrosine phosphorylation of VEGFR-2; (ii) the activation of downstream transductor phosphoinositide 3-OH kinase; and (iii) biological effects triggered by VEGF-A165. These results indicate a new role for alphavbeta3 integrin in the activation of an in vitro angiogenic program in endothelial cells. Besides being the most important survival system for nascent vessels by regulating cell adhesion to matrix, alphavbeta3 integrin participates in the full activation of VEGFR-2 triggered by VEGF-A, which is an important angiogenic inducer in tumors, inflammation and tissue regeneration.

Angiogenesis induced in vivo by hepatocyte growth factor is mediated by platelet-activating factor synthesis from macrophages.

This study shows that the neoangiogenesis induced by hepatocyte growth factor (HGF) was associated with a local synthesis of platelet-activating factor (PAF) and was inhibited by the specific PAF receptor antagonist WEB 2170 in a murine model in which matrigel was injected s.c. as a substratum for angiogenesis. The synthesis of PAF was concomitant with the early migration of endothelial cells and infiltration of MAC-1-positive macrophages. Infiltration of lymphocytes and polymorphonuclear leukocytes was never observed. In vitro studies demonstrated that mouse peritoneal macrophages, but not two murine microvascular endothelial cell lines or human and bovine endothelial cells from large vessels, synthesized PAF after stimulation with HGF. Furthermore, macrophages expressed the transcript of HGF receptor encoded by the MET proto-oncogene and migrated after HGF challenge. The binding of HGF to its receptor was followed by the activation of the receptor tyrosine kinase domain and phosphorylation of the beta subunit. Leukocyte depletion with 5-fluorouracil and anti-MAC-1 Ab added to matrigel prevented the infiltration of macrophages, the synthesis of PAF and the angiogenesis induced by HGF. PAF extracted and purified from mice challenged with HGF induced a rapid angiogenic response, inhibited by WEB 2170. These results suggest that the angiogenic effect of HGF in vivo is mediated, at least in part, by PAF synthesized from macrophages infiltrating the matrigel plug.

Activation of JAK2 in human vascular endothelial cells by granulocyte-macrophage colony-stimulating factor.

Besides the regulation of hematopoiesis, granulocyte-macrophage colony-stimulating factor (GM-CSF)induces the expression of a functional program in endothelial cells (ECs) related to angiogenesis and to their survival in the bone marrow microenvironment. ECs express specific GM-CSF high-affinity binding sites, which mediate the proliferative and migratory response. We now report that ECs express the alpha and beta subunits of GM-CSF receptor (GM-CSFR), and that GM-CSF is able to activate the Janus kinase (JAK)2, a member of the cytosolic tyrosine kinase family, which is known to mediate signals of several non-tyrosine kinase receptors. JAK2 tyrosine phosphorylation, as well as activation of its catalytic activity, is induced by subnanomolar concentrations of GM-CSF and occurs within 3 minutes of stimulation and persists at least for 10 minutes. The effect is specific as inferred by the lack of effect of heat-inactivated GM-CSF or neutralized by specific antibodies and by the finding that interleukin-5, which utilizes a specific alpha chain and the same beta chain of GM-CSFR, does not phosphorylate JAK2. Furthermore, we show that the amount of JAK2 physically associated with GM-CSFR beta chain is increased after GM-CSF stimulation and that GM-CSF triggers both beta chain and JAK2 tyrosine phosphorylation. Taken together, these results suggest that biologic activities of GM-CSF in vascular endothelium may, in part, be elicited by GM-CSFR-mediated JAK2 activation.

The angiogenesis induced by HIV-1 tat protein is mediated by the Flk-1/KDR receptor on vascular endothelial cells.

The HIV-1 Tat protein transactivates HIV, viral and some host cell genes. Tat can be released by infected cells and acts extracellularly in the microenvironment, regulating functions of immunocompetent and mesenchymal cells. One of the most striking effects of Tat is the induction of a functional program in vascular cells related to angiogenesis and inflammation (migration, proliferation and expression of plasminogen activator inhibitor-1 and E selectin). Tat induces growth of Kaposi's sarcoma (KS) spindle cells and is angiogenic in vivo and in transgenic mice10-12. We previously reported that Tat is a direct angiogenic factor and noted the Tat arginine- and lysine-rich sequence is similar to that of other potent angiogenic growth factors, such as vascular endothelial growth factor-A (VEGF-A). It is possible that Tat mimics one of these factors by interacting with its growth factor tyrosine kinase receptor. Here we demonstrate that Tat specifically binds and activates the Flk-1/kinase insert domain receptor (Flk-1/KDR), a VEGF-A tyrosine kinase receptor (for review see ref. 13), and that Tat-induced angiogenesis is blocked by agents blocking the Flk-1/KDR receptor. Endothelial cell stimulation by Tat occurs in the absence of activation of FLT-1, another VEGF-A tyrosine kinase receptor.

Platelet-activating factor (PAF) induces the early tyrosine phosphorylation of focal adhesion kinase (p125FAK) in human endothelial cells.

Platelet-activating factor (PAF) is a potent activator of angiogenesis and controls the motility and the shape of vascular endothelium. The mechanism(s) whereby PAF exerts its action are in part known. Here we report that the biological active (R)PAF enantiomer administrated to cultured endothelial cells induces the early phosphorylation in tyrosine residues of focal adhesion kinase (p125FAX) and paxillin, two molecules involved in the early signaling and cytoskeleton assembly in cells that undergo integrin-mediated adhesion or are challenged by neuropeptides or lysophosphatidic acid. The phenomenon is rapidly turned on, lasts for a few minutes and is adhesion-independent indicating that the chain of events induced by (R)PAF, including p125FAK activation, precedes adhesion. The inhibitory effect of WEB2086, a PAF receptor antagonist, and the lack of activity exerted by the (S)PAF enantiomer, indicate that (R)PAF-mediated p125FAK activation, is PAF receptor-dependent. Calphostin C, an inhibitor of protein kinase C blocks the effect of (R)PAF on p125FAK phosphorylation suggesting that protein kinase C activation is up-stream the activation of this tyrosine kinase. When endothelial cells are exposed to a substratum that allows adhesion and spreading. (R)PAF-stimulated cells, change their adhesive phenotype and start migrating. Inhibitors of tyrosine kinases, like 3-(1,4,-dihydroxytetralyl) methylen-2-oxindole and herbimycin A, reduce the cells migration, the transendothelial flux of albumin and the enhancement of p125FAK activity induced by (R)PAF. The observation that increased tyrosine phosphorylation of p125FAK and its ensuring association with focal adhesion occurs rapidly upon (R)PAF challenge indicates that this signaling molecule has a primary and independent role also in the signaling cascade initiated by (R)PAF.

Anti-tumor activity of cytokines against opportunistic vascular tumors in mice.

Polyoma middle T (PmT)-transformed endothelial cells may represent a unique murine model for human opportunistic vascular tumors. The present study was designed to evaluate the anti-tumor potential of a panel of 13 cytokines against murine PmT-transformed endothelial cells. Interferon gamma (IFNgamma) and transforming growth factor beta 1 (TGFbeta1) substantially decreased in a dose-dependent manner the proliferation of a panel of 6 PmT-transformed cell lines. IFNalpha and tumor necrosis factor alpha(TNFalpha) had marginal anti-proliferative activity, whereas other molecules (interleukins-1, -2, -4, -6 and -13, IFNbeta, leukemia inhibitory factor, oncostatin M, granulocyte-macrophage colony-stimulating factor) caused no growth inhibition. IFNgamma and TGFbeta1 were therefore selected for further analysis of their mechanism of action and in vivo relevance. IFNgamma and TGFbeta1 reduced the activity of phosphatidylinositol-3-kinase and the production of phosphatidylinositol 3,4-biphosphate, without modifying the tyrosine kinase(s) activity associated with PmT. IFNgamma and TGFbeta1 were also tested for their ability to modify the in vivo growth of the PmT-transformed endothelial cells H5V in syngeneic C57B1/6 mice. Treatment with IFNnu and TGFbeta1 significantly delayed tumor growth and increased survival time. In contrast, treatment with IFNalpha and TNFalpha failed to prolong survival. In nude mice, IFNgamma and TGFbeta1 had a transient effect on tumor growth but no effect on survival, suggesting a contribution of T cells to the in vivo anti-tumor activity of these cytokines.

Middle T antigen-transformed endothelial cells exhibit an increased activity of nitric oxide synthase.

Endothelioma cell lines transformed by polyoma virus middle T antigen (mTa) cause cavernous hemangiomas in syngeneic mice by recruitment of host cells. The production of nitric oxide (NO), as measured by nitrite and citrulline production, was significantly higher in mTa-transformed endothelial cells in comparison with nontransformed control cells. The maximal activity of NO synthase (NOS) was about 200-fold higher in cell lysates from the tEnd.1 endothelioma cell line than in lysates from nontransformed controls, whereas the affinity for arginine did not differ. The biochemical characterization of NOS and the study of mRNA transcripts indicate that tEnd.1 cells express both the inducible and the constitutive isoforms. NOS hyperactivity is not a simple consequence of cell transformation but needs a tissue-specific mTa expression. Since tEnd.1-conditioned medium induces NOS activity in normal endothelial cells, most likely NOS hyperactivity in endothelioma cells is attributable to the release of a soluble factor. This NOS-activating factor, which seems to be an anionic protein, could stimulate tEnd.1 cells to express NOS by an autocrine way. By the same mechanism, tEnd.1 cells could induce NOS in the neighboring endothelial cells, and NO release could play a role in the hemangioma development. Such hypothesis is confirmed by our in vivo experiments, showing that the administration of the NOS inhibitor L-canavanine to endothelioma-bearing mice significantly reduced both the volume and the relapse time of the tumor.

Involvement of a serine protease in the synthesis of platelet-activating factor by endothelial cells stimulated by tumor necrosis factor-alpha or interleukin-1 alpha.

It has been shown that production of platelet-activating factor (PAF, 1-O-alkyl-2-acetyl-sn-glycero-3-phosphocholine) by endothelial cells (EC) stimulated with tumor necrosis factor (TNF)-alpha and interleukin (IL)-1 alpha requires the synthesis of new proteins and is regulated by anti-proteinases. Here, we demonstrate that TNF-alpha and IL-1 alpha induce the expression by EC of a 34-kDa diisopropyl fluorophosphate-binding protein immunoprecipitated by an anti-human elastase antibody. This protein is released in the medium and cleaves the chromogenic substrate N-methoxysuccinyl- Ala-Ala-Pro-Val p-anilide, which is specific for elastase. The generation of this elastase-like protein seems to be important for the synthesis of PAF induced by TNF-alpha and IL-1 alpha, as suggested by the following observations: (a) it precedes the synthesis of PAF; (b) the inhibitors of serine protease and anti-human elastase antibody prevent the synthesis of PAF and the activation of 1-O-alkyl-2-lyso-glycerophosphocholine acetyl-CoA: acetyltransferase, which is a key enzyme of the PAF remodelling pathway; (c) elastase, at concentrations similar to that detectable in the medium of cytokine-activated EC, elicits a rapid synthesis of PAF by EC. High-performance liquid chromatography-tandem mass spectrometric analysis of bioactive PAF demonstrates that the molecular species produced after stimulation of EC with TNF-alpha, IL-1 alpha or elastase are similar, with a predominant synthesis of the alkyl species. These results indicate that TNF-alpha and IL-1 alpha stimulate the production of a serine protease which is critical in the activation of enzymes involved in PAF synthesis, suggesting the potential involvement of this mechanism in the regulation of EC functions.

Actions of molecules which regulate hemopoiesis on endothelial cells: memoirs of common ancestors?

The proliferation and differentiation of hematopoietic stem cells (hematopoiesis) takes place in close contact with stromal cells and matrix in bone marrow. Hematopoiesis requires cytokines, collectively termed colony stimulating factors (CSFs), which act on progenitor cell populations and induce their commitment to a specific lineage. For instance, leukemia, inhibitor factor and stem cell factor act on pluripotent cells and immature progenitors, granulocyte-macrophage colony stimulating factor (GM-CSF) acts at early stages of the development of myelomonocytic lineage, whereas granulocyte-colony stimulating factor (G-CSF) and macrophage-colony stimulating factor (M-CSF) act on more mature cells of the same lineage and are only required later during the differentiation of this cell lineage. A second important element for the hematopoietic process is the presence of extracellular matrix proteins, which bind CSFs and correctly present the molecules to specific receptors present on the surface of the progenitor cells. Finally, stromal cells (i.e. fibroblasts, endothelial cells and adipocytes) which support the growth of hematopoietic stem cells in vitro, are crucial for the production of CSFs and protein matrix and regulate the passage of mature cells from bone marrow to bloodstream. Idiopathic myelofibrosis is an example of the relevance of microenvironment in hematopoiesis. This disease is characterized by fibroblast and basement membrane accumulation, appearance of myofibroblasts and modification of the capillary network and provokes a bone marrow aplasia. In this article we review recent studies on the role of hemopoietic cytokines on stromal cells, in particular on endothelial cells, and propose a double role for CSFs in hematopoiesis: to induce the commitment of progenitor cells and to maintain the behavior of bone marrow endothelial cells.

Oncostatin M activates phosphatidylinositol-3-kinase in Kaposi's sarcoma cells.

Oncostatin M (OM) is a polypeptide cytokine that induces autocrine and paracrine effects on AIDS-Kaposi's sarcoma (KS) cells (Nair et al., Science, 255, 1430-1432, 1992; Miles et al., Science, 255, 1432-1434, 1992). The signalling pathways underlying this activation are largely unknown. We have found that OM binding to KS cell lines in vitro identifies a higher affinity binding site (Kd 10-20 pM) with a lower affinity (Kd 1.5 nM), high capacity binding site. The binding of OM to its receptor at the KS cell surface stimulates a rapid tyrosine phosphorylation of multiple proteins, including the p85 regulatory subunit of phosphatidylinositol-3-kinase (PI3K). In addition OM can stimulate the in vivo activation of PI3K and increases the PI3K activity in anti-phosphotyrosine and anti-src kinase family antibody directed immunoprecipitates. Genistein, an inhibitor of tyrosine kinases, inhibits the synthesis of phosphatidylinositol 3,4-biphosphate and the growth of KS cells. Finally, OM enhances tyrosine kinase activity in immune complex kinase assay performed with antibody anti-src kinase family. These data suggest that in KS cells OM can stimulate formation of tyrosine kinase co-ordinate signalling complexes, containing at least src kinase family and PI3K, which can drive the accumulation of the putative second-messengers D3-phosphorylated phosphoinositides.