Fibroblast growth factor 2 and the protease activity of tumor cells isolated from BK virus/tat transgenic mice

1996 ◽  
Vol 10 (5-6) ◽  
pp. 309-315 ◽  
Author(s):  
J.A.M. Maier ◽  
M. Mariotti ◽  
L. Meneghini ◽  
U. Cavallaro ◽  
Z. Wu ◽  
...  
2016 ◽  
Vol 117 (3) ◽  
pp. 599-611 ◽  
Author(s):  
Marja M. Hurley ◽  
Douglas J. Adams ◽  
Liping Wang ◽  
Xi Jiang ◽  
Patience Meo Burt ◽  
...  

Blood ◽  
2012 ◽  
Vol 120 (21) ◽  
pp. 3638-3638
Author(s):  
Rajendra Gharbaran ◽  
Andre Goy ◽  
Takemi Tanaka ◽  
Jongwhan Park ◽  
Christopher J Kim ◽  
...  

Abstract Abstract 3638 Although the vast majority of patients with Hodgkin Lymphoma (HL) do well with standard frontline therapy, about 20% will continue to relapse and have poor clinical outcome, even with high dose therapy and autologous stem cell transplant. Recent progress in the understanding of HL has led to an increasing awareness of the importance of the microenvironment and the cross-talk with cytokines and chemokines (JCO, Steidl et.al. 2011). However, given a low number of HL tumor cells present in lymph node biopsies, it is difficult to develop predictive biomarkers criteria for clinical outcome solely based on the characteristics of Reed Sternberg cells. To identify potential HL biomarkers, a bioinformatics approach was used to derive 151 genes that are potentially related to lymphomagenesis and outcome. These genes were screened against a library of non-tissue (mostly effusions) derived 9 HL cell lines that originated from refractory/relapsing HL patients. Two genes, Fibroblast Growth Factor-2 (FGF2/bFGF) and Syndecan-1 (SDC1/CD138), were found consistently overexpressed across all HL cell lines when compared to normal purified B-cells from PBL. We then tested a panel of lymphoma tissues including Hodgkin and non-Hodgkin subtypes by immunohistochemistry. FGF2 and SDC1 again were found overexpressed only in HL against 148 other lymphoma tissue samples (p<0.005). To evaluate the potential prognostic significance of these two genes for HL, we selected two groups of patients in the frontline setting; the first group is “good outcome” (n=12) with no relapse over 4 years and the second group is “bad outcome” (n=7) with multiple relapses/primary failures/deaths. When PBL collected at baseline, (IRB approved signed ICF) prior to chemotherapy was analyzed, FGF2 and SDC1 were overexpressed 25 fold (p<0.005) and 7150 fold (p<0.005) respectively in chemo-naive HL samples of the “bad outcome” versus “good outcome” patient subsets. In these PBL samples, the primary HL markers CD30 and CD15 were also upregulated by 70 fold (p<0.005) and 50 fold (p<0.005) respectively in the “bad outcome” versus “good outcome” groups, suggesting that those CD30+/CD15+ cells are potentially representing circulating HL tumor cells. This finding is further supported by significant downregulation of established markers that represent other common PBL cell subtypes, including CD4 (Helper T cells, −93 fold, p<0.005), CD8 (cytotoxic T cells, −6 fold, p<0.005), CD14/63 (monocytes, −467 fold (p<0.005) and −190 fold (p<0.005) respectively) and CD38 (plasma cells, −55 fold, p<0.005), CD19 (normal B cells, −200 fold with p<0.005) in the chemo-naive group with the “bad outcome” over the “good outcome” groups. These molecular signatures of immune cells and FGF2/SDC1 were not found in the PBL of chemo-exposed relapsed/refractory HL patients (n=8), suggesting that these expression signatures are limited to chemo-naive HL patients. To investigate FGF2 and SDC1 expression at tissue level, 67 archived HL samples (unrelated to the PBL samples) were analyzed by qRT-PCR and immunohistochemical methods. Tissues that fit the criteria of the bad outcome (n=9) group showed an intense immunostaining of FGF2 and SDC1 while the good outcome (n=20) group showed a moderate staining. The bad outcome group also showed significantly higher (65 fold, p<0.005) expression of CD68 mRNA over the good outcome group, indicating greater accumulation of cells of macrophage lineage as previously reported. FGF2 and SDC1 are overexpressed 25 fold (p<0.005) and 45 fold (p<0.005) respectively, in the bad outcome over the good outcome group when normalized against the normal lymph nodes, suggesting that only selected populations of CD30+ cells overexpress FGF2 and SDC1. When the relationship between the overexpression of FGF2/SDC1 and the metastatic potential was analyzed, the tissues derived from the “bad outcome” group showed very strong immunohistochemisty results for both MMP9 and TGF-β and overexpressed MMP9 (53 fold, p<0.005) and TGF-β (60 fold, p<0.005) mRNAs when compared to the “good outcome” group. Double immunofluorescent studies showed that CD30+ cells from the “bad outcome” group coexpressed FGF2 and SDC1 and that the same cells frequently coexpressed MMP9 and TGF-β suggesting that selected HL tumor cells may undergo shedding from the microenvironment and then be found as circulating tumor cells at baseline and correlate with very poor outcome. Disclosures: Mato: Celgene: Speakers Bureau; Millennium: Speakers Bureau; Seattle Genetics: Speakers Bureau; Genentech: Speakers Bureau.


2019 ◽  
Author(s):  
Cyril Legrand ◽  
Roberto Saleppico ◽  
Jana Sticht ◽  
Fabio Lolicato ◽  
Hans-Michael Müller ◽  
...  

SummaryFibroblast Growth Factor 2 (FGF2) is a tumor cell survival factor that is exported from cells by an unconventional secretory pathway. This process is based on direct translocation of FGF2 across the plasma membrane. FGF2 membrane translocation depends on PI(4,5)P2-induced formation of membrane-inserted FGF2 oligomers followed by extracellular trapping of FGF2 at the outer leaflet mediated by cell surface heparan sulfate proteoglycans. Beyond the well-characterized core mechanism of FGF2 membrane translocation, the Na,K-ATPase has been proposed to play a so far unknown role in unconventional secretion of FGF2. Here, we define a direct physical interaction of FGF2 with a subdomain of the cytoplasmic part of the α1 subunit of the Na,K-ATPase. Employing NMR spectroscopy and molecular dynamics simulations, we identified two lysine residues on the molecular surface of FGF2 that are shown to be essential for its interaction with α1. In intact cells, the corresponding lysine-to-glutamate variants of FGF2 were characterized by inefficient secretion and reduced recruitment to the inner plasma membrane leaflet as shown by single molecule TIRF microscopy. Our findings suggest that α1 acts upstream of PI(4,5)P2 facilitating efficient membrane translocation of FGF2 to the cell surface of tumor cells.


Sign in / Sign up

Export Citation Format

Share Document