Rap1 stabilizes beta-catenin and enhances beta-catenin-dependent transcription and invasion in squamous cell carcinoma of the head and neck

Abstract

Purpose: In head and neck squamous cell carcinoma (HNSCC) cells, Rap1 shuttles between the nucleus and cytoplasm. Prior findings suggested that Rap1 may modulate the beta-catenin-independent Wnt pathway in some settings, but the role of Rap1 in beta-catenin-dependent Wnt signaling remains undefined.

Experimental design and results: We observed that beta-catenin bound to active Rap1 in vitro and Rap1 activated beta-catenin/T-cell factor (TCF)-dependent transcription. Immunofluorescence studies showed that ectopic expression of Rap1 increased nuclear translocation of beta-catenin. Overexpression of active Rap1 facilitated an increase in beta-catenin-mediated transcription that was abrogated by dominant-negative TCF4. Conversely, small interfering RNA-mediated inhibition of endogenous Rap1 expression inhibited beta-catenin/TCF-mediated transcription as well as invasion of HNSCC. Furthermore, inhibition of Rap1 expression downregulated the expression of matrix metalloproteinase 7, a transcriptional target of beta-catenin/TCF. In HNSCC cells stably transfected with beta-catenin or treated with lithium chloride or Wnt3A to stabilize endogenous beta-catenin, inhibition of Rap1 expression led to decreases in the free pool of beta-catenin. Immunohistochemical studies of tissue from HNSCC patients revealed that increased beta-catenin intensity correlated with higher tumor stage. Furthermore, the prognostic effect of active Rap1 on tumor N stage was found to depend on cytosolic beta-catenin expression (P < 0.013). When beta-catenin is high, higher Rap1GTP intensity is associated with more advanced N stage.

Conclusions: The findings suggest that Rap1 enhances beta-catenin stability and nuclear localization. In addition to indicating that Rap1 has a significant role in regulating beta-catenin and beta-catenin-dependent progression to more advanced N-stage lesions, these data highlight Rap1 as a potential therapeutic target in HNSCC.

Figure 1

Cytoplasmic and nuclear pools of free β-catenin are detectable in human SCC. A. UM-SCC-(1, 14A, 17B, 22B, 74A, and 81B) cells grown on Lab-tek slides were incubated with β-catenin monoclonal antibody (1:3000) followed by secondary antibody, DAB detection, and counterstaining with hematoxylin. IgG control was negative (not shown). Cytosolic and nuclear staining (1, 74A; arrows), cytosolic, membrane and nuclear (17B, 81B; arrows) or membrane staining (22B; arrows) were observed. (14A, 74A and 81B, bar = 60 μm; 17B and 22B, bar = 40 μm; 1, bar = 20 μm). B. Whole-cell lysates prepared from DLD-1, UM-SCC-(1, 14A, 17B, 22B, 74A, 81B) and OSCC3 were immunoblotted with anti-β-catenin (upper panel) and GAPDH (upper middle panel), as a loading control. Free β-catenin was retrieved with the FLAG-tagged cytoplasmic tail of E-cadherin and immunoblotted with anti-β-catenin (lower middle panel) and anti-Flag (lower panel), as a loading control. C.β-catenin binds to Rap1_GTP. Cell lysates of human SCC cell lines UM-SCC-(14A, 17B, 22B, 81B, 1, 5) were evaluated for active Rap1 by the ralGDS pull down assay (upper left panel). The same filter was blotted with anti-β-catenin (lower left panel). Whole cell lysates were blotted to show total Rap1, β-catenin, and actin (middle panel). Control for ralGDS pull-down assay (right panel). For the ralGDS pull-down assay, beads were incubated alone or with whole cell lysates of 17B and 22B in the presence or absence of ralGDS as indicated (right panel, lanes 3–8). Total Rap1 in whole cell lysates from 17B and 22B was also immunoblotted (right panel, lanes 1 and 2). D. HEK293 cells were co-transfected with HA-tagged Rap1AG12 or pcDNA control vector and β-catenin as indicated. HA-Rap1G12 was immunoprecipitated (IP) and the immunoprecipitates were blotted with antiβ-catenin (upper left panel) and anti-HA (lower left panel) antibodies. Whole cell lysates were immunoblotted to show total β-catenin (upper right panel), HA (middle right panel), and GAPDH (lower right panel).

Figure 2

Rap1 enhances β-catenin/TCF mediated transcription. A. HEK293 cells were co-transfected with pTOPflash or pFOPflash β-catenin mediated reporter genes, Renilla luciferase, wild type β-catenin (Wt β-cat; +: 0.1 μg), the active form of Rap1A (Rap1AG12V, +: 0.1 μg), and active form of Rac1 (Rac1G12V, +: 0.1μg) in a 24-well plate. β-catenin mediated transcription was assayed 24 h later. B. HEK293 cells were co-transfected with pTOPflash or pFOPflash reporter genes, Renilla luciferase, wild-type β-catenin (Wt β-cat, +: 0.2μg, ++: 0.4μg, +++: 0.6μg), and Rap1A G12V (+: 0.9μg), as indicated, in a 6-well plate. Luciferase was assayed 24 h later (*, p<0.05). C, HEK 293 cells were co-transfected with a dominant negative mutant form of TCF4 (Mut TCF4, +: 0.1μg) and pTOPflash, pFOPflash, Renilla luciferase, wild type β-catenin (Wt β-cat, +: 0.1 μg), and Rap1A G12V (+: 0.1 μg), as indicated. The data are the average of at least three independent experiments (*, p<0.05).

Figure 3

Effects of Rap1 on β-catenin and cell invasion in vitro. A. Rap1 enhances nuclear translocation of β-catenin. HEK 293 cells were transfected with FLAG-tagged wild-type β-catenin (left upper three panels) or co-transfected with wild type β-catenin and Rap1AG12V (left lower three panels). Cells were fixed with methanol and incubated with anti-FLAG antibody (FITC) to detect FLAG-tagged β-catenin (green fluorescence). The slides were counterstained with DAPI nuclear stain (x40 magnification). Quantification of FITC-labeled β-catenin in the nucleus or cytoplasm of 293 cells co-transfected with β-catenin and pcDNA or β-catenin and Rap1A G12 (*, p<0.05) (right panel). Data are representative of two independent experiments. B. SCC cells stably transfected wth HA-tagged β-catenin or empty vector were transfected with HA-tagged Rap1A G12V or pcDNA. Expression of Rap1A G12V and β-catenin were verified by immunoblot analysis with anti-HA antibody as well as β-catenin and Rap1 antibodies. GAPDH was used as a loading control. C. Free β-catenin was retrieved with the cytoplasmic domain of E-cadherin and evaluated by immunoblot analysis (upper panel). Free β-catenin, quantified by densitometry, was normalized to FLAG and expressed as percent of control (lower panel). D. Rap1 promotes β-catenin-mediated invasion in SCC cells. UM-SCC-1 cells stably transfected with HA-tagged β-catenin or empty vector were transfected with HA-tagged Rap1A G12V or pcDNA. Transfected cells suspended in DMEM were added to the upper chamber of a matrigel insert and DMEM/FBS was introduced in the lower chamber. After overnight incubation, the invasive cells were stained and quantified. Data from two independent experiments are shown.

Figure 4

Rap1A-β-catenin mediated transcription and invasion of SCC. A. UM-SCC-1 cells stably transfected with HA-tagged β-catenin or empty vector were nucleofected with siRNA Rap1A or Non-target siRNA. Cell lysates were immunoblotted with Rap1, HA, FLAG, GAPDH, β-catenin and MMP7 antibodies. Free β-catenin was retrieved with the cytoplasmic domain of E-cadherin. B. UM-SCC-1 cells stably transfected with HA-tagged β-catenin or empty vector were nucleofected with siRNA Rap1A or Non-target siRNA concurrently with pTOPflash, pFOPflash, and Renilla luciferase. Luciferase activity was assayed 24h later. Data are representative of three independent experiments, each in triplicate. C. UM-SCC-1 cells stably transfected with HA-tagged β-catenin or empty vector were nucleofected with siRNA Rap1A or Non-target siRNA. Transfected cells suspended in DMEM were added to the upper chamber of a matrigel insert and DMEM/FBS was introduced in the lower chamber. After overnight incubation, the invasive cells were stained and quantified. Data are representative of three independent experiments. D. UM-SCC-1 cells nucleofected with siRNA Rap1A or non target siRNA were treated with lithium chloride (40 mM, 6h; two independent experiments, each in duplicate) or Wnt3A (50 ng/ml, 6h; one experiment in duplicate) as indicated. Free β-catenin and whole cell lysates were immunoblotted with Flag, Rap1, β-catenin and GAPDH antibodies.

Figure 5

The prognostic effect of active Rap1 on tumor N-stage depends on the expression of cytosolic β-catenin. A. A 5μm tissue section of a human HNSCC tissue microarray was incubated with β-catenin antibody (left panel) or IgG control (not shown) or Rap1 antibody (right panel) as described in the Methods section. After DAB reaction, the slides were counterstained with hematoxylin (bars = 50 μm). High cytosolic β-catenin and rap1GTP staining are shown for the same tumor specimen. B. High β-catenin intensity is associated with more advanced T-stage as assessed by the Spearman Correlation Coefficient and represented in box and whisker plots. Patient groups: T1, tumor size <2 cm (n = 4); T2, tumor size 2–4 cm (n = 10); T3, tumor size >4cm (n = 16), T4, tumor invades through cortical bone, inferior alveolar nerve, floor of mouth or skin of face (n = 13). C. The proportional odds model was used to assess the association between N stage and the interaction of Rap1 and β-catenin. When cytosolic β-catenin was high, higher active Rap1 intensity was prognostic of more advanced N-stage; when β-catenin intensity is low, lower active Rap1 is associated with more advanced N-stage. Early N-stage is N0 (red triangles, n=6) and N1 (green triangles, n=6); advanced N-stage is N2 (blue triangles, n=15) and N3 (purple triangles, n=3).

Figure 6

Proposed Model for interaction between Rap1, β-catenin and MMP7 in SCC progression. Rap1 promotes invasion via β-catenin mediated MMP7 secretion and inhibits invasion via inhibition of MMP9 and MMP2 secretion. Tumors with high β-catenin expression are associated with more advanced tumor stage. Rap1 promotes invasion via β-catenin mediated effects and inhibits invasion via inhibition of MMP9 and MMP2.