|
|
Blood, Vol. 112, Issue 6, 2463-2473, September 15, 2008
Oncogenic Kit controls neoplastic mast cell growth through a Stat5/PI3-kinase signaling cascade
Blood Harir et al.
112: 2463
Supplemental materials for: Harir et al
Files in this Data Supplement:
- Figure S1. cS5F-transduction induces mast cell hyperplasia in transplanted mice (JPG, 139 KB)
-
(A) Shown above is a detailed view of typical organic alterations of gastrointestinal tracts of vGFP-transplanted (n=4) and cS5F-transplanted mice (n=4) sacrificed 5 weeks post-transplant as confirmed with individual transplants. Leukemic cS5F-transplanted mice had a unique pathologic appearance with enlarged liver, splenomegaly, mesenterial lymphadenopathy (yellow plus symbol) and with severe alterations of the gastrointestinal tract (black plus symbol). Stomachs of mice transplanted with vGFP appear normal but cS5F–transplanted mice display severe inflammation in the stomach (red plus symbol). Stomachs of diseased mice did not contain significant amounts of food in contrast to the vGFP transplanted mice. (B) Immunohistochemical analysis of large bowel mucosa and submucosa of cS5F-transplanted versus vGFP-transplanted mice. Consecutive tissue sections were stained for P-Y-Stat5 and tryptase with specific antibodies (400× magnification). H&E staining was used to observe organotypic structure changes.
- Figure S2. Nuclear versus cytoplasmic localization of P-Y-Stat5 in growth hormone (GH) stimulated liver controls and leukemic cells in stomach submucosa sections of cS5F-transplanted mice (JPG, 113 KB)
-
Control P-Y-Stat5 immunostaining was performed on liver sections of WT mice injected either with PBS or with growth hormone, and treated as described previously (Engblom et al., Genes Dev. 2007) (A–B; 400× magnification). In contrast to cS5F-transplanted mice, tissues of growth hormone-stimulated mice display P-Y-Stat5 reactivity predominantly in the nucleus (B). Immunohistochemical detection of P-Y-Stat5 in stomach submucosa and mucosa sections of cS5F-transplanted mice is shown in the lower panels. Antigen retrieval was performed either with 10 mM citrate buffer, pH 6.0 (C–D; 400× and 630× magnification, respectively) or CC1 buffer, pH 8.0 (E–F; 400× magnification;Ventana Medical Systems). The P-Y-Stat5 immunostaining was performed in the presence or absence of a blocking peptide (C–D). With either of the antigen retrieval protocols P-Y-Stat5 staining was detected predominantly in the cytoplasm. Irrespective of the rather weak staining, the high pH antigen retrieval method could confirm the strong cytoplasmic staining obtained with the citrate buffer retrieval protocol.
- Figure S3. Automated recognition and analysis of immunohistochemical images with HistoQuest (JPG, 133 KB)
-
(A) HistoQuest is a cytometry-based application for automated recognition of cells in immunohistochemical images. The input is any regular chromogen-stained image in RGB format, as provided by the color camera attached to the microscope. (B) Results are displayed in a scattergram, where 2 color parameters are plotted against each other in order to visualize cell populations based on cut-off values that divide the measured objects in “positive”and “negative”populations. The cut-off values for background staining can be chosen manually utilizing the forward/backward gating tool of the HistoQuest software. The latter links each cell in the captured image to a single dot in the scattergram and enables exact differentiation of “positive” and “negative” cells. This kind of data representation is well-known from flow-cytometry. Each single dot represents a single nucleus/cell within the tissue context and is recorded as follows: (C) and (E) After interactive color definition of the nuclear staining (e.g. hemalaun) and the specific antibody reactivity (e.g. anti-tryptase), the colors in the RGB image are separated and shown in virtual color channels. Similarly to color channels in fluorescence microscopy the virtual color channels are shown as grey value images. Color separation is performed by an proprietary algorithm (www.tissuegnostics.com). (D) and (F) Based on the color-separated images the individual nuclei and the corresponding chromogen patterns per cell are segmented. The resulting labeled binary mask (where each color represents an individual object) is used as measure mask for determination of several parameters like object size and chromogen density. All parameters are stored in a database for each object identified.
- Figure S4. cS5F-induced mast cell development is independent of autocrine expression and secretion of IL-3 (JPG, 74.9 KB)
-
(A) Sorted GFP+-bone marrow cells from cS5F-grafted mice (cS5F-BM) or from control GFPv-grafted mice (GFPv-BM) were cultured in the presence of SCF alone (10 ng/ml) or with IL-3 (10 ng/ml). Cells were counted daily using the trypan blue dye exclusion method. In contrast to cS5F-BM cells, growth of control GFPv-BM cells required the presence of SCF and IL-3. (B) Toluidine blue staining of cS5F-BMMC cells and control GFPv-BMMC obtained after 2 and 3 weeks in culture respectively in presence of SCF (10 ng/ml) alone (cS5F-BMMC) or SCF and IL-3 (10 ng/ml) (GFPv-BMMC). (C) RT-PCR reactions were performed with RNAs from cS5F-BM cells isolated from two different leukemic mice (cS5F-BM1 and 3), an IL-3-dependent Ba/F3 cell line and the IL-3-expressing cell line Wehi-3B with specific primers for murine IL-3, OSM and GAPDH as control. Results showed that none of these cells express IL-3 mRNAs while IL-3 expression was detected in the IL-3-producer cell line Wehi-3B. In contrast, expression of the Stat5 target gene oncostatin m (osm) was clearly observed in cS5F-BM1 and cS5F-BM3 cells or in Ba/F3 and Wehi-3B cell lines that retain Stat5 activation upon IL-3 stimulation. (D) cS5F-BM cells (cS5F-BM1 and 3) and Ba/F3 cells were incubated or not with neutralizing antibodies raised against murine IL-3 (1 g/ml). The number of viable cells was determined daily using the Trypan Blue dye exclusion method. Addition of anti-IL-3 antibodies did not block the growth of cS5F-BM cells but clearly inhibited the growth of the IL-3-dependent Ba/F3 cell line. (E–F) SCF activates Stat5 in human but not in murine mast cells. Control GFPv-BMMC obtained after three weeks in culture in presence of SCF and IL-3 were deprived of IL-3 and SCF for six hours and then re-stimulated with SCF or IL-3 during 30 min and 1 hr. Cell extracts were then analyzed by Western blot with the indicated antibodies. SCF-induced tyrosine phosphorylation of Stat5 was also evaluated in the SCF-dependent human mast cell line LAD-2. Extracts from LAD-2 cells that were first deprived during six hours, left untreated or treated for 30 min with SCF or IL-3, were analyzed by Western blot with the indicated antibodies.
- Figure S5. Transcriptional activity of P-Y-Stat5 proteins in HMC-1 cells (JPG, 48.5 KB)
-
(A) Nuclear extracts were analysed in a band shift experiment with a Stat5 binding site from the β-casein gene promoter as probe. Extracts were incubated with anti-Stat5a (S5A) or anti-Stat5b (S5B) antibodies and the supershifts (SS) were monitored. (B) HMC-1 cells were transiently transfected with a (Stat5RE)x6-tk-luciferase reporter construct containing six copies of the Stat5 binding site. As control, cells were also transfected with the empty tk-luciferase reporter construct. The following day, luciferase activities were determined in cell extracts. (C–D) Transduction of a TAT-Stat5 recombinant protein containing 2 point mutations (TAT-Stat5EE437;438AA) in the DNA binding domain (DBD) partially inhibits HMC-1 cell growth. (C) Schematic representation of the TAT-Stat5 constructs. The different cDNAs were introduced into the bacterial expression vector pTAT-HA. Resulting recombinant Stat5 (wt and (EE437;438AA) proteins were fused in their N-terminal part to a 6xHis-Tag followed by the protein transduction domain (PTD) of the TAT protein and a HA tag sequence. The purification was performed as previously described (Harir et al., 2007). (D) HMC-1 cells were transduced with 10 nM of each recombinant proteins or not (NaCl). The number of living cells was determined six days later. Results are the mean of three independent experiments.
- Figure S6. Effects of LY294002 and Pimozide on Stat5 and Akt phosphorylation in HMC-1 cells (JPG, 51 KB)
-
HMC-1 cells were treated or not with the PI 3-kinase inhibitor LY294002 or pimozide at the indicated concentrations. The levels of P-Y-Stat5 and P-S-Akt proteins were next determined by Western blot with the indicated antibodies. Results showed that LY294002 blocks Akt but not Stat5 activation while pimozide inhibits the phosphorylation of both molecules.
- Figure S7. Gab2 interacts with Stat5 in HMC-1 cells and is involved in the growth of these neoplastic mast cells (JPG, 58.5 KB)
-
(A) Gab2 was immunoprecipitated from HMC-1 cell lysate (BM1 and BM2). Subsequently, the content in p85 and Stat5 proteins was evaluated in the Gab2 immunoprecipitates by Western blotting using the indicated antisera. Cell lysates were also immunoprecipitated with an isotypic control IgG antibody. (B) Gab2 was first immunodepleted or not from HMC-1 cell lysates prior to immunoprecipitation with anti-Stat5 antibodies. The presence Gab2 in the Stat5 immunoprecipitates was next evaluated by Western blot with an anti-Gab2 antibody. (C) Schematic representation of the TAT-Gab2 constructs. Purification of TAT-Gab2 proteins was performed as previously described (Harir et al., Blood 2007). (D) HMC-1 cells were transduced or not with 100 nM of the different TAT-Gab2 proteins during 24 hours and the presence of the recombinant proteins was determined by Western blot with anti-HA antibodies. (E) HMC-1 cells were transduced or not with 100 nM of the different TAT-Gab2 proteins and the number of living cells was determined every three days using the trypan blue dye assay. Results showed that expression of a Gab2 mutant deficient in p85 binding (Gab2-3YF) inhibits the growth of HMC-1 cells.
- Figure S8. Tetramer deficient cS5F mutants activate Akt in Ba/F3 cells (JPG, 54 KB)
-
(A) cS5F and the tetramer deficient mutants cS5FW37A and cS5F136 cDNAs were cloned into the bi-cistronic expression vector pIREShrGFP carrying a Flag tag sequence.The resulting constructs and the empty vector (mock) were introduced in Ba/F3 cells by electroporation. The following day, GFP+ cells were sorted by flow cytometry and grown in absence or presence of IL-3 for 16 hours. Cell extracts were next prepared and analyzed by Western blot with the indicated antibodies. (B) Cytoplasmic localization of cS5F and the tetramer-deficient cS5F mutants in transfected Ba/F3 cells. In each case, cells were spun on cytospin slides and analyzed by indirect immunocytochemistry using an anti-P-Y-Stat5 antibody (AX1). (C) Sorted GFP+ cells expressing cS5F or the tetramer-deficient cS5F mutants were grown in absence of IL-3 and the number of living cells were daily enumerated using the Trypan Blue dye exclusion method. Results shown are representative of three different transfection assays.
|
|
