Sphingosine-1-phosphate receptor-3 signaling up-regulates epidermal growth factor receptor and enhances epidermal growth factor receptor-mediated carcinogenic activities in cultured lung adenocarcinoma cells

Abstract

Sphingosine-1-phosphate (S1P) regulates a wide array of biological functions. However, the role of S1P signaling in tumorigenesis remains to be elucidated. In this study, we show that S1P receptor subtype 3 (S1P₃) is markedly up-regulated in a subset of lung adenocarcinoma cells compared to normal lung epithelial cells. Specific knockdown of S1P₃ receptors inhibits proliferation and anchorage-independent growth of lung adenocarcinoma cells. Mechanistically, we demonstrate that S1P₃ signaling increases epidermal growth factor receptor (EGFR) expression via the Rho kinase (ROCK) pathway in lung adenocarcinoma cells. Nuclear run-off analysis indicates that S1P/S1P₃ signaling transcriptionally increases EGFR expression. Knockdown of S1P₃ receptors diminishes the S1P-stimulated EGFR expression in lung adenocarcinoma cells. Moreover, S1P treatment greatly enhances EGF-stimulated colony formation, proliferation and invasion of lung adenocarcinoma cells. Together, these results suggest that the enhanced S1P₃-EGFR signaling axis may contribute to the tumorigenesis or progression of lung adenocarcinomas.

Figure 1

Increased expression of S1P₃ in cultured human lung adenocarcinoma cells. (A) Expression levels of S1P receptors in normal SAEC (primary culture of normal human small airway epithelial cells) and A549 adenocarcinoma cells (left panel), as well as LLC cells (right panel). -ve, PCR were performed without cDNA template. Note that S1P₃ receptors are markedly increased in cancerous lung cells, compared to normal lung epithelial cells. (B) Real-time PCR quantitation of S1P₃ expression in 3 immortalized normal lung epithelial and 9 lung adenocarcinoma cells lines. Note that S1P₃ is elevated, ranging from 2–18 fold increases, in lung adenocarcinoma cells compared to immortalized normal lung epithelial cells. 3-KT, HBEC3-KT; 2-E, HBEC2-E; 2-KT, HBEC2-KT. MCF human breast cancer cells were used as a positive control. (C) RT-PCR analysis for the expression of S1P receptor subtypes in normal lung epithelial and lung carcinoma cells. h3bB4, the constantly expressed acidic ribosomal phosphoprotein P0 was used as the loading control.

Figure 2

Knockdown of S1P₃ diminishes lung adenocarcinoma cell proliferation and colony formation. (A) H1793 human lung adenocarcinoma cells were stably transfected with si-S1P₃ or pRS control vector. The expression of S1P receptor subtypes was measured by semi-qunatitative RT-PCR (upper panel) and real-time PCR (lower panel). Note that S1P₃ receptors were profoundly knocked-down in si-S1P₃ stably transfected cells. The knockdown is specific since si-S1P₃ transfection had no effect on the expression of GAPDH (upper panel) or S1P_{1, 2, 4, 5} receptors (lower panel). Real-time PCR data (mean ± SD of triplicate determinations. ^**p<0.01, t-test) are shown by 2^−ΔCt values (ΔCt = Ct of S1P receptor - Ct of GAPDH) to represent the relative abundance of S1P receptor subtypes in H1793 cells. (B) Cell proliferation in control vector (pRS) and si-S1P₃ transfected H1793 cells. Data represent mean ± SD of four determinants. The knockdown of S1P₃ receptors markedly reduced cell proliferation. ^*p<0.05, t-test. (C) Colony formation assay. 100 (upper panels) and 500 (lower panels) cells were seeded for clonogenic assay. Numbers in parenthesis are mean ± SD of colony numbers (n=4; ^*p<0.05, t-test).

Figure 3

S1P transcriptionally activates EGFR expression via ROCK pathway in lung adenocarcinoma cells. (A) HBEC2-KT and H1793 cells were stimulated with or without S1P (300 nM) for 4 h. The expression of indicated genes was measured by real-time PCR. Note that S1P stimulation significantly induced EGFR in H1793 cells, whereas S1P was unable to induce EGFR expression in HBEC2-KT cells. Data are mean ± SD (n=3). CDH1, E-cadherin; TLR, toll-like receptor; F2R, thrombin receptor. ^*p<0.01 (t-test). (B) Nuclear run-off analysis indicates that S1P activates EGFR transcription. Nuclei were isolated from H1793 with or without S1P treatment (300 nM) for 4 h. Nuclear run-off assays were performed as we previously described (28). Upper panel, RT-PCR; lower panel, real-time PCR (mean ± SD, n=3). ^*p<0.05 (t-test). (C) H1793 were pre-treated for 1 h with or without pharmacological inhibitors of various signaling molecules, followed by stimulating with S1P (200 nM) for 4 h. Inhibitors used are: SP600125 for JNK, SB203580 for p38 kinase, Bay 11–7085 for NFκB, LY294002 for PI3-kinase, and Y-27632 for Rho kinase (ROCK). Data are mean ± SD of triplicate determinations. ^*p<0.01, t-test.

Figure 4

S1P-enhanced EGFR expression is mediated by S1P₃ receptors. (A) H1793 and HBEC2-KT cells were treated with S1P (300 nM) for various times. The expression of EGFR was quantitated by real-time PCR analysis. Note that S1P induces EGFR expression in a time-dependent manner in H1793 cells, whereas S1P was unable to enhance EGFR in HBEC2-KT cells. Data are mean ± SD of triplicate determinations. ^*p<0.05, vs. control untreated cells, t-test. (B) H1793 cells were treated with indicated concentrations of S1P for 4 h. Note that S1P dose-dependently induced EGFR expression. ^*p<0.05, vs. control untreated cells, t-test (n=3). (C) H1793 cells were treated with S1P (300 nM) for indicated times. EGFR polypeptides were detected by Western blotting with anti-EGFR. The nitrocellulose membrane was reprobed with β-actin antibody to show the loading control. Lower panel, the intensity of immunoreactive EGFR band was quantified by a densitometer, normalized to β-actin, and represent as fold increase (n=4). ^*p<0.05, vs. control untreated cells, t-test. (D) Treatment of VPC23019 inhibits S1P-induced EGFR up-regulation. H1793 were stimulated with or without S1P (300 nM), in the presence or absence of VPC23019 (5 μM) for 4 h. ^*p<0.05 (t-test, n=3). (E) Knockdown of S1P₃ abolishes S1P-induced EGFR expression. H1793 cells were stably transfected with si-S1P₃ or pRS control vector as described in the Materials and methods section. Following S1P stimulation (300 nM, 4 h), the expression of EGFR (upper panel) and S1P₃ (lower panel) was measured by real-time PCR. ^*p<0.05, vs. untreated cells, t-test (n=3). (F) S1P induced EGFR expression in four other lung carcinoma cell lines; however, S1P was unable to induce EGFR in the normal HBEC3-KT lung epithelial cells. ^*p<0.01 (t-test, n=3).

Figure 5

S1P treatment potentiates EGF on proliferation, anchorage-independent growth, and invasion of lung adenocarcinoma cells. H1793 (A) and A549 (B) (8,000 cells in 200 μl) were resuspended in plain RPMI medium containing indicated concentrations of S1P and/or EGF for three days. Cell proliferation was measured as described in the Materials and methods section. Note that neither S1P nor EGF stimulated cell proliferation in this concentration range, whereas S1P plus EGF significantly stimulated cell proliferation (^*p<0.01, t-test). Data are mean ± SD (n=3) of a representative experiment, which were repeated three times with similar results. (C) Clonogenic assays were performed as described in Materials and methods section. Note that treatment of S1P and EGF together significantly induced colony formation compared to treatment of S1P alone (n=3; ^*p<0.05, t-test). Cell invasion was measured in H1793 (D) and A549 (E) cells treated with indicated concentrations of S1P and/or EGF. Note that EGF-induced cell invasion was greatly enhanced in the presence of S1P (n=6; ^*p<0.05, t-test). (F) Invasive capability of H1793 cells was measured in the indicated combinations of treatments. Note that Cay10444 (5 μM) completely inhibited, whereas gefitinib (0.5 μM) reduced approximately 40% of the S1P (100 nM) and EGF (0.05 ng/ml)-induced H1793 invasion. Data are mean ± SD (n=3) of a representative experiment, which was repeated two times with similar results.