REGULATION OF HGF EXPRESSION BY ΔEGFR-MEDIATED C-MET ACTIVATION IN GLIOBLASTOMA CELLS

Overexpression of the hepatocyte growth factor receptor (c-Met) and its ligand, the hepatocyte growth factor (HGF), and a constitutively active mutant of the epidermal growth factor receptor (∆EGFR/EGFRvIII), occur frequently in glioblastoma. c-Met is activated in a ligand-dependent manner by HGF or in a ligand-independent manner by ∆EGFR. Dysregulated c-Met signaling contributes to the aggressive phenotype of glioblastoma, yet the mechanisms underlying the production of HGF in glioblastoma are poorly understood. We found a positive correlation between HGF and c-Met expression in glioblastoma, suggesting that they are coregulated. This is supported by the finding that in a c-Met/HGF axis-dependent glioblastoma cell line, shRNA-mediated silencing of c-Met, or treatment with the c-Met inhibitor SU11274, attenuated HGF expression. Biologically, c-Met knockdown decreased anchorage-independent colony formation and the tumorigenicity of intracranial xenografts. Building on prior findings that ∆EGFR enhanced c-Met activation, we found that ∆EGFR also led to increased HGF expression, which was reversed upon ∆EGFR inhibition with AG1478. ∆EGFR required c-Met to maintain elevated HGF expression, colony formation of glioblastoma cells, and the tumorigenicity of orthotopic xenografts. An unbiased mass spectrometry-based approach identified phosphotyrosine-related signaling changes that occurred with c-Met knockdown in a glioblastoma cell line expressing ΔEGFR and in parental cells. Notably, phosphorylation of STAT3, a master regulator of the mesenchymal GBM subtype and a known target of ∆EGFR, also decreased when c-Met was silenced in these cells, suggesting that the signals from these receptors converge on STAT3. Using a STAT3 inhibitor, WP1193, we showed that STAT3 inhibition decreased HGF mRNA expression in ΔEGFR-expressing glioblastoma cells. Consistent with these findings, constitutively active STAT3 partially restored HGF expression and anchorage-independent growth of c-Met knockdown glioblastoma cells that overexpressed ΔEGFR. We found that higher levels of HGF and c-Met expression associated with the mesenchymal GBM subtype. Taken together, these results suggest that the activity of c-Met regulates the expression of HGF in glioblastoma cells, that ∆EGFR feeds positively into this autocrine loop, that signaling of the two receptors together modulate HGF expression via STAT3, and that the HGF/c-Met axis may therefore be a good additional target for therapy of mesenchymal GBM tumors.

Transcriptional analysis of liver-specific expression and differential interleukin-6 response of rat kininogen genes

Analyses of rat T1 kininogen gene/chloramphenicol acetyltransferase (T1K/CAT) constructs revealed two regions important for tissue-specific and induced regulation of T1 kininogen.^ Although the T1 kininogen gene is inducible by inflammatory cytokines, a highly homologous K kininogen gene is minimally responsive. Moreover, the basal expression of a KK/CAT construct was 5- to 7-fold higher than that of the analogous T1K/CAT construct. To examine the molecular basis of this differential regulation, a series of promoter swapping experiments was carried out. Our transfection results showed that at least two regions in the K kininogen gene are important for its high basal expression: a distal 19-bp region (C box) constituted a binding site for CCAAT/enhancer binding protein (C/EBP) family proteins and a proximal 66-bp region contained two adjacent binding sites for hepatocyte nuclear factor-3 (HNF-3). The distal HNF-3 binding site from the K kininogen promoter demonstrated a stronger affinity than that from the T1 kininogen promoter. Since C/EBP and HNF-3 are highly enriched in the liver and known to enhance transcription of liver-specific genes, differential binding affinities of these factors accounted for the higher basal expression of the K kininogen gene.^ In contrast to the K kininogen C box, the T1 kininogen C box does not bind C/EBP presumably due to their two-nucleotide divergence. This sequence divergence, however, converts it to a consensus binding sequence for two IL-6-inducible transcription factors--IL-6 response element binding protein and acute-phase response factor. To functionally determine whether C box sequences are important for their differential acute-phase response, T1 and K kininogen C boxes were swapped and analyzed after transfection into Hep3B cells. Our results showed that the T1 kininogen C box is indeed one of the IL-6 response elements in T1 kininogen promoter. Furthermore, its function can be modulated by a 5$\sp\prime$-adjacent C/EBP-binding site (B box) whose mutation significantly reduced the overall induced activity. Moreover, this B box is the target site for binding and transactivation of another IL-6 inducible transcription factor C/EBP$\delta.$ Evolutionary divergence of a few critical nucleotides can either lead to subtle changes in the binding affinities of a given transcription factor or convert a binding sequence for a constitutive factor to a site recognized by an inducible factor. (Abstract shortened by UMI.) ^

Cytochrome P450 4F isoforms in different species: Identification, gene regulation and putative roles in inflammation

The cytochrome P450 4F subfamily comprises a group of enzymes that metabolize derivatives of arachidonic acid such as prostaglandins, lipoxins leukotrienes and hydroxyeicosatetraenoic acids, which are important mediators involved in the inflammatory response. Therefore, we speculate that CYP4Fs might be able to modulate the extent of the inflammation by controlling of the tissue levels of these inflammatory mediators, especially, leukotriene B4. One way to provide support for this hypothesis is to test whether the expression of CYP4Fs changes under inflammatory conditions, since these changes are required to adjust the levels of inflammatory mediators. ^ A lipopolysacchride (LPS) induced rat inflammation model was used to analyze the expressions of rat CYP4F4 and CYP4F5 in liver and kidney. LPS administration did not change the constitutive expression level of CYP4F4 and CYP4F5. In liver, the expressions of CYP4F4 and CYP4F5 decreased to 50–60% of the untreated level. The same effect of LPS on CYP4F4 and CYP4F5 expression can be mimicked in hepatocyte primary cultures treated with LPS, indicating a direct of effect of LPS on hepatocytes. LPS treatment also decreased the activity of liver microsomes towards chlorpromazine, however, antibody inhibition study revealed that liver CYP4Fs are not the only players in metabolizing chlorpromazine. To study further the underlying mechanism, CYP4F5 gene was isolated, characterized, and the promoter region was defined. ^ Accumulating evidence showed that peroxisome proliferator-activated receptors (PPARs) play an active role in inflammation. To investigate the possible role of PPARα in regulating CYP4F expression by inflammation or by clofibrate treatment, the expressions of two new mouse 4F isoforms were analyzed in PPARα knockout mice upon LPS or clofibrate challenge. A novel induction of CYP4F15 by LPS and clofibrate was observed in kidney, and this effect is totally dependent on the presence of PPARα. Renal CYP4F16 expression was not affected by LPS or clofibrate in both (+/+) and (−/−) mice. In contrast, hepatic expressions of CYP4F15 and CYP4F16 were reduced significantly in (+/+) mice, but much less in (−/−) mice, suggesting that PPARα is partially responsible for this down-regulation. Clofibrate treatment reduced the expression of CYP4F16 in liver, but has no effect on CYP4F15 and PPARα does not have a role in hepatic CYP4F expression regulated by clofibrate. In general, CYP4Fs are regulated in an isoform-, tissue- and species-specific manner. ^ A human CYP4F isoform, CYP4F11, was isolated. The genomic structure was also solved by using database mining and bioinformatics tools. Localization of CYP4F11 to chromosome 19, 16 kb upstream of CYP4F2, suggests that human CYP4F genes may form a cluster on chromosome 19. This novel human 4F is highly expressed in liver, as well as in kidney, heart and skeletal muscle. Further study of the activity and gene regulation on CYP4F11 will provide us more insights into the physiological functions of CYP4F subfamily. ^

The heparan sulfate-fibroblast growth factor signaling system in liver growth and function

The heparan sulfate (HS)-fibroblast growth factor (FGF) signaling system is a ubiquitous regulator that senses local environmental changes and mediates cell-to-cell communication. This system consists of three mutually interactive components. These are regulatory polypeptides (FGF), FGF receptor (FGFR) and heparan sulfate proteoglycans (FGFRHS). All four FGFR genes are expressed in the adult liver. Expression of the FGFR1–3 genes is generally associated with non-parenchymal cells while expression of the FGFR4 gene is associated with parenchymal hepatocytes. We showed that livers of mice lacking FGFR4 exhibited normal morphology and regenerated normally in response to partial hepatectomy. However, the FGFR4 (−/−) mice exhibited depleted gallbladders, an elevated bile acid pool and elevated excretion of bile acids. Cholesterol- and bile acid-controlled liver cholesterol 7α-hydroxylase (Cyp7a), the limiting enzyme for bile acid synthesis, was elevated, unresponsive to dietary cholesterol, but repressed normally by dietary cholate. These results indicated that FGFR4 was not directly involved in liver growth but exerted negative control on liver bile acid synthesis. This was confirmed in transgenic mice overexpressing the constitutively active human FGFR4 in livers. The transgenic mice exhibited decreased fecal bile acid excretion, bile acid pool size, and expression of Cyp7a. Introduction of this constitutively active human FGFR4 into FGFR4 (−/−) mice restored the inhibition of bile acid synthesis. Activation of the c-Jun N-terminal Kinase (JNK) pathway by FGFR4 correlated with the repressive effect on bile acid synthesis. ^ To determine whether FGFR4 played a broader role in liver-specific metabolic function, we examined the impact of both acute and chronic exposure to CCl 4 in FGFR4 (−/−) mice. Following acute CCl4 exposure, the FGFR4 (−/−) mice exhibited accelerated liver injury, a significant increase in liver mass and delayed hepatolobular repair, with no apparent effect on liver cell proliferation and restoration of cellularity. Chronic CCl4 exposure resulted in severe fibrosis in livers of FGFR4 (−/−) mice compared to normal mice. Analysis at both mRNA and protein levels indicated an 8 hr delay in FGFR4-deficient mice in the down-regulation of cytochrome P450 2E1 (CYP2E1) protein, the major enzyme whose products underlie CCl 4-induced injury. These results show that hepatocyte FGFR4 protects against acute and chronic insult to the liver and prevents accompanying fibrosis. ^ Of the 23 FGF polypeptides, FGF1 and FGF2 are present at significant levels in the liver. To determine whether FGF1 and FGF2 played a role in CCl 4-induced liver injury and fibrosis, we examined the impact of both acute and chronic exposure to CCl4 in both wild-type and FGF1-FGF2 double-knockout mice. Following acute CCl4 exposure, FGF1(−/−)FGF2(−/−) mice exhibited accelerated liver injury, overall normal liver growth and repair, and decreased liver collagen α1(I) induction. Liver fibrosis resulting from chronic CCl4 exposure was markedly decreased in livers of FGF1(−/−)FGF2(−/−) mice compared to wild-type mice. This study suggests a role for FGF1 and FGF2 in hepatic fibrogenesis. ^ In summary, our three part study shows that specific components of the ubiquitous HS-FGF signaling family in the liver context interfaces with metabolite- and xenobiotic-controlled networks to regulate liver function, but has no apparent direct effect on liver cell growth. ^

The survival role and the potential therapeutic targeting of MET in multiple myeloma

Multiple myeloma (MM) is a debilitating and incurable B-cell malignancy. Previous studies have documented that the hepatocyte growth factor (HGF) plays a role in the pathobiology of MM. The receptor tyrosine kinase MET induced signaling initiates when its ligand HGF binds to the MET receptor. However, the direct importance of MET in MM has not been elucidated. The present work used three different but complementary approaches to reduce MET protein levels or its activity to demonstrate the importance of MET in MM. ^ In the first approach, MET transcript and protein levels were reduced by directly targeting the cellular MET transcripts using shRNA retroviral infection techniques. This direct reduction of MET mRNA leads to a reduction of MET protein levels, which caused an inhibition of growth and induction of cell death. ^ In the second approach, a global transcription inhibitor flavopiridol was used as a potential pharmacological tool to reduce MET levels. MET has a short half-life of 30 min for mRNA and 4 hours for protein; therefore using a RNA pol II inhibitor such as flavopiridol would be a viable option to reduce MET levels. When using flavopiridol in MM cell lines, there was a reduction of MET transcript and protein levels, which was associated with the induction of cell death. ^ Finally in the last strategy, MET kinase activity was suppressed by MP470, a small molecule inhibitor that binds to the ATP binding pocket in the kinase domain. At concentrations where phosphorylation of MET was inhibited there was induction of cell death in MM cell lines and primary cells from patients. In addition, in MM cell lines there was a decrease in phosphorylation of AKT (ser473) and caspase-9 (ser196); downstream of MET, suggesting that the mechanism of action for survival may be through these cascade of events. ^ Overall, this study provides a proof-of-principle that MET is important for the survival of MM cell lines as well as primary plasma cells obtained from patients. Therefore, targeting MET therapeutically may be a possible strategy to treat patients with this debilitating disease of MM. ^

Analysis of Wnt/beta-catenin signaling in kidney tubulogenesis

The β-catenin/Lef/Tcf-mediated Wnt pathway is central to the developmental of all animals, stem cell renewal, and cancer progression. Prior studies in frogs and mice have indicated that the ligand Wnt-4 is essential for the mesenchyme to epithelial transition that generates tubules in the context of kidney organogenesis. More recently, Wnt-9b in mice, was likewise found to be required. Yet despite the importance of Wnt signals in renal development, the corresponding Frizzled receptor(s) and downstream signaling mechanim(s) are unclear. My work addresses these knowledge gaps using in vitro (Madin-Darby Canine Kidney cells) and in vivo (Xenopus laevis and zebrafish pronephros) tubulogenic kidney model systems. Employing established reporter constructs of Wnt/β-catenin pathway activity, I have determined that MDCK cells are highly responsive to Wnt-4, -1, and -3A, but not to Wnt-5A and control conditions. I have confirmed that Wnt-4's canonical signaling activity in MDCK cells is mediated by downstream effectors of the Wnt/β-catenin pathway using β-Engrailed and dnTCF-4, constructs that suppress this pathway. I have further found that MDCK cells express the Frizzled-6 receptor, and that Wnt-4 forms a biochemical complex with Frizzled-6, yet does not appear to transduce Wnt-4's canonical signal. Additionally, I demonstrate that standard Hepatocyte Growth Factor (HGF)-mediated (non-physiologic) induction of MDCK tubulogenesis in collagen matrices is not altered by activation or suppression of β-catenin signaling activity; however, β-catenin signaling maintains cell survival in this in vitro system. Using a Wnt/β-catenin signaling reporter in Xenopus laevis, I detect β-catenin signaling activity in the early pronephric epithelial kidney tubules. By inhibiting the Wnt/β-catenin signaling pathway in both zebrafish and Xenopus , a significant loss of kidney tubulogenesis is observed with little or no effect on adjoining axis or somite development. This inhibition also leads to the appearance of severe edema that phenocopies embryos depleted for Wnt-4. Tubulogenic loss does not appear to be caused by increased cell death in the Xenopus pronephric field, but rather by lessened expression of tubule epithelium genes associated with cellular differentiation. Together, my results show that Wnt/β-catenin signaling is required for renal tubule development and that Wnt-4 is a strong candidate for activating this pathway. ^

Intercellular adhesion and membrane polarity in rat hepatocytes: Localization of cell -CAM 105 and other domain-specific membrane proteins {\it in situ\/} and {\it in vitro\/} by electron microscopy

Cell-CAM 105 has been identified as a cell adhesion molecule (CAM) based on the ability of monospecific and monovalent anti-cell-CAM 105 antibodies to inhibit the reaggregation of rat hepatocytes. Although one would expect to find CAMs concentrated in the lateral membrane domain where adhesive interactions predominate, immunofluorescence analysis of rat liver frozen sections revealed that cell-CAM 105 was present exclusively in the bile canalicular (BC) domain of the hepatocyte. To more precisely define the in situ localization of cell-CAM 105, immunoperoxidase and electron microscopy were used to analyze intact and mechanically dissociated fixed liver tissue. Results indicate that although cell-CAM 105 is apparently restricted to the BC domain in situ, it can be detected in the pericanalicular region of the lateral membranes when accessibility to lateral membranes is provided by mechanical dissociation. In contrast, when hepatocytes were labeled following incubation in vitro under conditions used during adhesion assays, cell-CAM 105 had redistributed to all areas of the plasma membrane. Immunofluorescence analysis of primary hepatocyte cultures revealed that cell-CAM 105 and two other BC proteins were localized in discrete domains reminscent of BC while cell-CAM 105 was also present in regions of intercellular contact. These results indicate that the distribution of cell-CAM 105 under the experimental conditions used for cell adhesion assays differs from that in situ and raises the possibility that its adhesive function may be modulated by its cell surface distribution. The implications of these and other findings are discussed with regard to a model for BC formation.^ Analysis of molecular events involved in BC formation would be accelerated if an in vitro model system were available. Although BC formation in culture has previously been observed, repolarization of cell-CAM 105 and two other domain-specific membrane proteins was incomplete. Since DMSO had been used by Isom et al. to maintain liver-specific gene expression in vitro, the effect of this differentiation system on the polarity of these membrane proteins was examined. Based on findings presented here, DMSO apparently prolongs the expression and facilitates polarization of hepatocyte membrane proteins in vitro. ^

Characterization of the cholesterol 7$\alpha$-hydroxylase promoter in hypercholesterolemia -resistant rabbits

Our laboratory has developed and partially characterized a strain of New Zealand white rabbits that are resistant to the hypercholesterolemia which typically occurs in normal rabbits when fed a cholesterol-enriched diet. This phenotype is most likely attributed to an increase in bile acid excretion by hypercholesterolemia-resistant (CRT) rabbits as a result of elevated enzyme activity of cholesterol 7$\alpha$-hydroxylase (C7$\alpha$H), the rate-limiting enzyme in bile acid synthesis. Northern analysis revealed that CRT rabbits, in comparison to normal rabbits, have a 7-fold greater steady-state C7$\alpha$H mRNA levels irrespective of dietary regimen. The C7$\alpha$H gene in both phenotypes was determined to be a single copy gene. The hypothesis was that the elevated C7$\alpha$H mRNA levels in CRT rabbits, in comparison to normal animals, was due to an increase in the transcription rate of the C7$\alpha$H gene as a result of a mutation in a cis-acting element and/or a trans-acting factor within the hepatocyte. To isolate the C7$\alpha$H gene from both normal and CRT rabbits, genomic libraries were prepared from both phenotypes into $\lambda$GEM12 vectors using conventional techniques. Three CRT and one normal phage clones that contained the C7$\alpha$H gene were identified by screening the library with a series of probes located within different exons of the C7$\alpha$H cDNA. Sequencing analysis confirmed that approximately 1100 bp of the C7$\alpha$H 5'-flanking region from both normal and CRT phenotypes was identical. The increase in C7$\alpha$H mRNA levels was not attributed to a cis-acting mutation within this region. Liver nuclear extracts were prepared from normal and CRT rabbits maintained either on a basal or 0.25% cholesterol-enriched diet and incubated with several radiolabeled DNA fragments from the C7$\alpha$H gene. A 37 basepair region, located between nucleotides $-$452 to $-$416 was identified that had altered binding patterns between normal and CRT rabbits as a function of diet. Two additional regions, $-$747 to $-$575 and $-$580 to $-$442, produced banding patterns which were identical, irrespective of phenotype or diet. In conclusion, these studies suggested that the increase in C7$\alpha$H mRNA in CRT rabbits was due to differences in binding of a cholesterol-responsive transcription factor to the C7$\alpha$H promoter. ^

Identification, purification, sequencing and characterization of human lung fibroblast motility -stimulating factor for human soft tissue sarcoma cells

Paracrine motogenic factors, including motility cytokines and extracellular matrix molecules secreted by normal cells, can stimulate metastatic cell invasion. For extracellular matrix molecules, both the intact molecules and the degradative products may exhibit these activities, which in some cases are not shared by the intact molecules. We found that human peritumoral and lung fibroblasts secrete motility-stimulating activity for several recently established human sarcoma cell strains. The motility of lung metastasis-derived human SYN-1 sarcoma cells was preferentially stimulated by human lung and peritumoral fibroblast motility-stimulating factors (FMSFs). FMSFs were nondialyzable, susceptible to trypsin, and sensitive to dithiothreitol. Cycloheximide inhibited accumulation of FMSF activity in conditioned medium; however, addition of cycloheximide to the migration assay did not significantly affect motility-stimulating activity. Purified hepatocyte growth factor/scatter factor (HGF/SF), rabbit anti-hHGF, and RT-PCR analysis of peritumoral and lung fibroblast HGF/SF mRNA expression indicated that FMSF activity was unrelated to HGF/SF. Partial purification of FMSF by gel exclusion chromatography revealed several peaks of activity, suggesting multiple FMSF molecules or complexes.^ We purified the fibroblast motility-stimulating factor from human lung fibroblast-conditioned medium to apparent homogeneity by sequential heparin affinity chromatography and DEAE anion exchange chromatography. Lysylendopeptidase C digestion of FMSF and sequencing of peptides purified by reverse phase HPLC after digestion identified it as an N-terminal fragment of human fibronectin. Purified FMSF stimulated predominantly chemotaxis but chemokinesis as well of SYN-1 sarcoma cells and was chemotactic for a variety of human sarcoma cells, including fibrosarcoma, leiomyosarcoma, liposarcoma, synovial sarcoma and neurofibrosarcoma cells. The motility-stimulating activity present in HLF-CM was completely eliminated by either neutralization or immunodepletion with a rabbit anti-human-fibronectin antibody, thus further confirming that the fibronectin fragment was the FMSF responsible for the motility stimulation of human soft tissue sarcoma cells. Since human soft tissue sarcomas have a distinctive hematogenous metastatic pattern (predominantly lung), FMSF may play a role in this process. ^

Activation of c -Met contributes to growth and metastasis of human colorectal carcinoma

c-Met is the protein tyrosine kinase receptor for hepatocyte growth factor/scatter factor (HGF/SF) and mediates several normal cellular functions including proliferation, survival, and migration. Overexpression of c-Met correlates with progression and metastasis of human colorectal carcinoma (CRC). The goals of this study were to determine if overexpression of c-Met directly contributes to tumorigenicity and liver metastatic potential of colon cancer, and what are the critical pathways regulated by c-Met in this process. The studies used two colon tumor cell lines, KM12SM and KM20, which express high levels of constitutively active c-Met and are highly metastatic in nude mice. To examine the effects of c-Met overexpression, subclones of theses lines with reduced c-Met expression were obtained following transfection with a c-Met specific targeting ribozyme. Reduction of c-Met in KM12SM cells abolished liver metastases when cells were injected intrasplenically in an experimental metastasis assay. However, c-Met downregulation in theses clones was unstable. Three stable KM20 clones with a 25–35% reduction in c-Met protein levels but 60–90% reduction in basal c-Met autophosphorylation and kinase activity were obtained. While HGF increased c-Met kinase activity in the clones with reduced c-Met, the activity was less than that observed in parental or control transfected cells. Correlating with the reduction in c-Met kinase activity, subclones with reduced c-Met expression had significantly reduced in vitro growth rates, soft-agar colony forming abilities, and increased apoptosis. HGF/SF treatment did not affect anchorage-dependent growth or soft-agar colony forming abilities. Further, c-Met downregulation significantly impaired the ability of HGF/SF to induce migration. To examine the effects of reduced c-Met on tumor formation, parental and c-Met reduced KM20 cells were grown subcutaneously and intrahepatically in nude mice. c-Met downregulation delayed, but did not abolish growth at the subcutaneous site. When these cells were injected intrahepatically, both tumor incidences and size were significantly reduced. To further understand the molecular basis of c-Met in promoting tumor growth, the activation of several signaling intermediates that have been implicated in c-Met mediated growth, survival and migration were compared between KM20 parental cells and subclones with reduced c-Met expression levels. The expression and activity (as determined by phosphorylation) of AKT and Erk1/2 were unaltered. In contrast, Src kinase activity, as measured by immune complex kinase assay, was reduced 2–5 fold following c-Met downregulation. As Src has been implicated in growth, survival and migration, Src activation in c-Met overexpressing lines is likely contributing to the tumorigenic and metastatic capabilities of colon tumor cell lines that overexpress c-Met. Collectively, these results suggest that c-Met overexpression plays a causal role in the development of CRC liver metastases, and that c-Src and c-Met inhibitors may be of potential therapeutic benefit for late-stage colon cancer. ^