Lipid phosphate phosphatase-3 regulates tumor growth via β-catenin and CYCLIN-D1 signaling

Abstract

Background: The acquisition of proliferative and invasive phenotypes is considered a hallmark of neoplastic transformation; however, the underlying mechanisms are less well known. Lipid phosphate phosphatase-3 (LPP3) not only catalyzes the dephosphorylation of the bioactive lipid sphingosine-1-phosphate (S1P) to generate sphingosine but also may regulate embryonic development and angiogenesis via the Wnt pathway. The goal of this study was to determine the role of LPP3 in tumor cells.

Results: We observed increased expression of LPP3 in glioblastoma primary tumors and in U87 and U118 glioblastoma cell lines. We demonstrate that LPP3-knockdown inhibited both U87 and U118 glioblastoma cell proliferation in culture and tumor growth in xenograft assays. Biochemical experiments provided evidence that LPP3-knockdown reduced β-catenin, CYCLIN-D1, and CD133 expression, with a concomitant increase in phosphorylated β-catenin. In a converse experiment, the forced expression of LPP3 in human colon tumor (SW480) cells potentiated tumor growth via increased β-catenin stability and CYCLIN-D1 synthesis. In contrast, elevated expression of LPP3 had no tumorigenic effects on primary cells.

Conclusions: These results demonstrate for the first time an unexpected role of LPP3 in regulating glioblastoma progression by amplifying β-catenin and CYCLIN-D1 activities.

Figure 1

Expression of LPP3 in primary tumors. (A) Western blot was probed with an anti-LPP3-C-cyto antibody. Each lane contained 10 μg total protein prepared from primary human tumor tissues. LPP3 protein is indicated by two black arrows. (B) Tumor line cell extracts were analyzed by WB with the indicated antibodies. All blots shown are representative of those obtained from at least three separate experiments with similar results.

Figure 2

Immunohistochemistry of primary tumors. (A) A representative serial section of human glioblastoma was stained with hematoxylin and eosin (H&E), white arrows indicate microvascular proliferation, and (B) black arrows indicate anti-LPP3-RGD positive glioblastoma cells. Scale bars, 100 μm. (C) Anti-LPP3 staining in normal pancreas show cytosolic expression in both endocrine (Islet of Langerhans-arrow) and exocrine components. (D, E) In pancreatic endocrine tumors, 65% of the tumors show cytosolic staining. Inset: Nuclear localization of LPP3 is also detected in 40% of cases. (F) An example of LPP3-negative tumor cells. All experiments are representative of those obtained in at least three separate experiments, with similar results.

Figure 3

LPP3 regulates glioblastoma cell proliferation. (A) The time line indicates when glioblastoma cell lines were plated, shRNA (retrovirus) mediated knockdown was conducted, cells were growth factor and serum starved, fresh medium was added, BrdU (1.0 μg/ml) was added, and the proliferation assay was performed. (B) Quantification of glioblastoma cell proliferation assay. The data represent the mean ± s.e.m. n = 5-7 from five to seven independent experiments, ¥ P < 0.05 vs. control untreated; ¶ P < 0.05 vs. control shRNA treated group. (C-F) Representative images of the BrdU assay. Arrows indicate BrdU positive cells. Scale bar, 150 μm. (G) The efficiency of knockdown of LPP3 was assessed by western blotting with the indicated antibodies. Figure 3G (top panel, last lane, anti-LPP3), a partial band is likely produced by over flow from the neighboring well. All experiments are representative of those obtained in at least three separate experiments with similar results.

Figure 4

LPP3 regulates glioblastoma cell migration. (A) The time line indicates when glioblastoma cell lines were plated, shRNA (retrovirus) mediated knockdown was conducted, medium was added, and the migration assay was performed as described in materials and methods. (B) Quantification of glioblastoma cell migration through the Boyden chamber. The data represent the mean ± s.e.m. n = 5-7, ¥ P < 0.05 vs. control shRNA treated group. (C) Efficiency of knockdown of LPP3 was assessed by western blotting with the indicated antibodies. All experiments are representative of those obtained in at least three separate experiments with similar results.

Figure 5

LPP3 regulates glioblastoma tumor growth. (A) Stable clones of U87 and U118 were selected and efficiency of knockdown was determined by western blotting with the indicated antibodies. LPP3-knockdown decreased total β-catenin, CYCLIN-D1, and CD133 proteins, whereas it had no effect on LPP2 or GAPDH. (B) Conditioned media were collected and subjected to ELISA for VEGF, IL-8, and TIMP-2. LPP3-knockdown in U87 and U118 cells reduced the concentration of VEGF and IL-8, whereas TIMP-2 was unaffected. (C) In the xenograft assay, mice receiving cells that expressed control shRNA showed increased tumor growth (n = 12, ¥ P < 0.05) after 14 and 28 days. LPP3-knockdown reduced U87 and U118 tumor growth (n = 12, ¶ P < 0.05). (D) Representative images from the xenograft assay. Dotted circles indicate the location of tumor implants.

Figure 6

Forced expression of LPP3 potentiates SW480 tumor growth in the xenograft assay. (A) pLNCX2-based retroviral constructs expressing vector alone, hLPP1, hLPP2, hLPP3 or mLpp3 cDNAs. (B) The comparable levels of hLPP1, hLPP2, hLPP3 and mLpp3 proteins were confirmed by western blotting with anti-HA antibody. (C) Xenograft assay: there was minimal tumor growth after 14 and 28 days in mice receiving LPP3-deficient SW480 cells (2 × 10⁴ tumor cells/site) that expressed vector alone, hLPP1, or hLPP2 constructs. In contrast, tumor volume was significantly increased in mice receiving SW480 cells that expressed hLPP3 or mLpp3 (n = 12, ¥ P < 0.01 vs. hLPP1 group; n = 12, ¶ P < 0.005 vs. hLPP1 group) at 14 and 28 days.

Figure 7

Domains involved in the regulation of SW480 tumor growth. (A) Schematics of pLNCX2-based retroviral constructs (a-e). Empty red and blue circles indicate the relative positions of lipid phosphatase and RGD cell adhesion domains, respectively. Filled red and blue circles indicate phosphatase-inactive (PI) and RAD (adhesion defective) domains, respectively. Black boxes represent transmembrane segments. Three copies of hemagglutinin (HA) epitopes were fused to the N-terminal and in-frame to the open reading frame of the cDNAs. (B) Equivalent protein expression and phosphorylation states of β-catenin and CYCLIN-D1 proteins were analyzed by immunoblotting with anti-HA, anti-β-catenin, anti-p-β-catenin, and anti-CYCLIN-D1 antibodies. Equal loading of proteins across the lanes were determined with the anti-GRB-2 antibody. All blots are representative of those obtained in at least three separate experiments with similar results. (C) LPP3 expression potentiates SW480 tumor growth. Nude mice were injected subcutaneously with SW480 cells (~ 2 × 10⁴) expressing the indicated constructs. After 21 days, the tumor outgrowths were measured and photographed. (n = 12, ¶ P < 0.005 vs. vector alone group). (D) Representative pictures of primary tumors in nude mice injected with SW480 cells expressing the indicated constructs.