Expression of activated STAT5 in neoplastic mast cells in systemic mastocytosis: subcellular distribution and role of the transforming oncoprotein KIT D816V

Abstract

Recent data suggest that the signal transducer and activator of transcription (STAT)5 contributes to differentiation and growth of mast cells. It has also been described that constitutively phosphorylated STAT5 (pSTAT5) plays a pro-oncogenic role in various myeloid neoplasms. We examined the expression of pSTAT5 in neoplastic mast cells in systemic mastocytosis and asked whether the disease-related oncoprotein KIT D816V is involved in STAT5 activation. As assessed by immunohistochemistry using the anti-pSTAT5 antibody AX1, neoplastic mast cells were found to display pSTAT5 in all SM patients examined (n = 40). Expression of pSTAT5 was also demonstrable in the KIT D816V-positive mast cell leukemia cell line HMC-1. Using various staining-protocols, pSTAT5 was found to be located in both the cytoplasmic and nuclear compartment of mast cells. To define the functional role of KIT D816V in STAT5-activation, Ba/F3 cells with doxycycline-inducible expression of KIT D816V were used. In these cells, induction of KIT D816V resulted in an increased expression of pSTAT5 without substantial increase in total STAT5. Moreover, the KIT D816V-targeting kinase-inhibitor PKC412 was found to counteract expression of pSTAT5 in HMC-1 cells as well as doxycycline-induced expression of pSTAT5 in Ba/F3 cells. Finally, a dominant negative STAT5-construct was found to inhibit growth of HMC-1 cells. Together, our data show that neoplastic mast cells express cytoplasmic and nuclear pSTAT5, that KIT D816V promotes STAT5-activation, and that STAT5-activation contributes to growth of neoplastic mast cells.

Figure 1

Immunohistochemical detection of phosphorylated (p) STAT5 in neoplastic mast cells in the bone marrow of patients with systemic mastocytosis (SM). Serial sections prepared from paraffin-embedded bone marrow (iliac crest) of a patient with indolent SM (A–F) and a patient with aggressive SM (G–L) were stained with anti-tryptase antibody G3 (A, D, G, J), and the anti-pSTAT5 antibody AX1 using two staining protocols, one ′predominantly cytoplasmic′ staining protocol (B, E, H, K), and a second ′predominantly nuclear′ staining protocol (C, F, I, L). Serial section-staining revealed that almost all of the tryptase-positive neoplastic mast cells co-expresssed cytoplasmic pSTAT5, and many of these cells also exhibited nuclear pSTAT5 (C, F, I, L). The lower panels (D, E, F, and J, K, L) represent higher magnifications (×100) of the respective upper panels (A, B, C, and G, H, I) (×40). In separate experiments, AX1 was pre-incubated with control buffer (M, O) or a STAT5-specific blocking peptide (N, P) before being applied, which resulted in a negative stain in both the predominantly cytoplasmic staining protocol (M, N) and predominantly nuclear staining protocol (O,P).

Figure 2

Detection of phosphorylated (p) STAT5 in the cytoplasm of primary neoplastic mast cells and in the mast cell leukemia cell line HMC-1. Bone marrow mast cells obtained from a patient with mast cell leukemia (A, B) and HMC-1 cells (C, D) were spun on cytospin slides and then were stained with the anti-pSTAT5 antibody AX1 (A, C). Wright-Giemsa staining confirmed the presence of immature mast cells (B, D). E, F: KIT D816V-negative HMC-1.1 cells (E) and KIT D816V+ HMC-1.2 cells (F) were stained with serial dilutions of AX1 antibody as indicated using two different staining protocol, ie, one conventional protocol (protocol A: left panels in E and F), and one protocol for optimal visualization of nuclear pSTAT5 (protocol B: right panels in E and F). In both staining protocols, the reactivity of the cytoplamic compartment (black bars) and nuclear compartment (open bars) with AX1 antibody was determined separately. Results are expressed as percentage of reactive cells in one typical experiment (the same results were obtained in a second independent experiment). G–J: The AX1 antibody was pre-incubated with control buffer (G, I) or a STAT5-specific blocking peptide (H, J) for 1 hour, and then was applied to HMC-1 cells using two different staining protocol, ie, one conventional staining protocol (G, H), and one for optimal visualization of nuclear pSTAT5 (I, J). K, L: Flow cytometry was performed with an Alexa488-conjugated anti-pSTAT5 antibody (shaded plot) or an isotype-matched control antibody (unshaded plot). As visible, the antibody was found to react with pSTAT5 in KIT D816V-positive HMC-1.2 cells (K), as well as in KIT D816V-negative HMC-1.1 cells (L).

Figure 3

Quantitative analysis of cytoplasmic and nuclear pSTAT5 in HCM-1 cells. Expression of pSTAT5 in cytoplasmic extracts (CE) and nuclear extracts (NE) of HMC-1 cells was determined in three independent experiments (#1, #2, #3) by Western blotting. For control purpose, extracts were also examined for expression of Raf-1 (cytoplamic marker) and topoisomerase-1 (nuclear marker antigen). A shows results from Western blotting, and B provides a densitometric evaluation of the data obtained with cytoplamic extracts (open bars) and nuclear extracts (black bars) in the three experiments (#1, #2, #3).

Figure 4

Induction of STAT5 activity by KIT D816V in Ba/F3 cells. STAT DNA-binding acitivity was analyzed using extracts of unstimulated Ton.Kit.D816V cells (co, without KIT D816V) and extracts of Ton.Kit.D816V cells induced to express KIT D816V by exposure to doxycycline (+Doxycycline). Before being exposed to doxycycline, cells were incubated with control medium or medium containing imatinib (1 μmol/L) or midostaurin (1 μmol/L) at 37°C for 4 hours. Cell extracts were analyzed using blunt-ended annealed oligonucleotides. For STAT5 analysis in electrophoretic mobility shift assays (EMSA), the proximal STAT-binding element of the bovine β-casein promoter was used. Binding reactions were performed by incubating the radiolabeled probe with cell-lysates (20 μg) for 30 minutes. In supershift-reactions of STAT-containing complexes, 2 μg of antibodies specific for the C-terminal transactivation-domains of STAT1, STAT3, and STAT5 were added before EMSA was performed. Samples were separated by electrophoresis. As shown, the KIT D816V-induced phosphorylation of STAT5 was inhibited by PKC412/midostaurin.

Figure 5

Effects of KIT TK inhibitors on pSTAT5 expression in HMC-1 cells. A: Western blot analysis of KIT D816V-positive HMC-1.2 cells using antibodies against pSTAT5. Before immunoprecipitation (IP) and Western blotting, cells were incubated in control medium, imatinib (1 μmol/L), midostaurin/PKC412 (1 μmol/L), or nilotinib/AMN107 (1 μmol/L) at 37°C for 4 hours. Western blotting and IP were performed as described in the text. IP was conducted using a polyclonal anti-STAT5a antibody. Western blotting was performed using the anti-phospho-tyr antibody 4G10 for detection of pSTAT5, and anti-STAT5a antibody for detection of total STAT5. B–D: Flow cytometric assessement of expression of pSTAT5 in HMC-1 cells. B, C: Cells were incubated with control medium (open graphs) or with KIT inhibitors (gray graphs), ie, AMN107/nilotinib, 1 μmol/L (B) and PKC412/midostaurin, 1 μmol/L (C) at 37°C for 4 hours. D: Dose-dependent effect of PKC412 on expression of pSTAT5 in HMC-1 cells. Cells were incubated in control medium (co) or various concentrations of PKC412 as indicated (4 hours). Then, cells were permeabilized by methanol and subjected to flow cytometry using an antibody to pSTAT5 as described in the Materials and Methods.

Figure 6

Effect of dominant negative (dn) STAT5 on growth of HMC-1 cells. HMC-1 cells were transduced with a dn STAT5 construct (MSCV-STAT5BΔ754-IRES-GFP) (gray bars) or with a GFP-labeled vector control (black bars) as described in the text. Then, mixtures of transfected and nontransfected HMC-1 cells were prepared and cultured for various time periods. The number (percentage) of GFP-positive cells (relative to all viable cells) was determined by flow cytometry after various time intervals as indicated. Results represent the mean ± SD of three experiments performed in parallel.

Figure 7

Effects of piceatannol on growth and viability of HMC-1 cells. HMC-1.1 cells (lacking KIT D816V) and HMC-1.2 cells (expressing KIT D816V) were incubated in control medium (co) or with various concentrations of piceatannol as indicated at 37°C for 48 hours (A) or 24 hours (B, C). A: To determine cell growth, cultured cells were subjected to MTT assay as described in the text. Results are expressed as percentage of control and represent the mean ± SD of four independent experiments. B: After incubation with piceatannol, the numbers (percentage) of apoptotic cells were determined by light microscopy. Results show the percentage of apoptotic cells and represent the mean ± SD of three independent experiments. C: Numbers (percentage) of active caspase 3-positive HMC-1.2 cells after incubation with various concentrations of piceatannol as indicated. Expression of active caspase 3 was determined by immunocytochemistry as described in the text. Results show the percentage of caspase 3-positive cells and represent the mean ± SD of three independent experiments. *P < 0.05.