Ovarian cancer-derived lysophosphatidic acid stimulates secretion of VEGF and stromal cell-derived factor-1 alpha from human mesenchymal stem cells

Abstract

Lysophosphatidic acid (LPA) stimulates growth and invasion of ovarian cancer cells and tumor angiogenesis. Cancer-derived LPA induces differentiation of human adipose tissue-derived mesenchymal stem cells (hASCs) to alpha-smooth muscle actin (alpha-SMA)-positive cancer-associated fibroblasts. Presently, we explored whether cancer-derived LPA regulates secretion of pro-angiogenic factors from hASCs. Conditioned medium (CM) from the OVCAR-3 and SKOV3 ovarian cancer cell lines stimulated secretion angiogenic factors such as stromal-derived factor-1 alpha (SDF-1 alpha) and VEGF from hASCs. Pretreatment with the LPA receptor inhibitor Ki16425 or short hairpin RNA lentiviral silencing of the LPA((1)) receptor abrogated the cancer CM-stimulated expression of alpha-SMA, SDF-1, and VEGF from hASCs. LPA induced expression of myocardin and myocardin-related transcription factor-A, transcription factors involved in smooth muscle differentiation, in hASCs. siRNA-mediated depletion of endogenous myocardin and MRTF-A abrogated the expression of alpha-SMA, but not SDF-1 and VEGF. LPA activated RhoA in hASCs and pretreatment with the Rho kinase inhibitor Y27632 completely abrogated the LPA-induced expression of alpha-SMA, SDF-1, and VEGF in hASCs. Moreover, LPA-induced alpha-SMA expression was abrogated by treatment with the ERK inhibitor U0126 or the phosphoinositide-3-kinase inhibitor LY294002, but not the PLC inhibitor U73122. LPA-induced VEGF secretion was inhibited by LY294002, whereas LPA-induced SDF-1 secretion was markedly attenuated by U0126, U73122, and LY294002. These results suggest that cancer-secreted LPA induces differentiation of hASCs to cancer-associated fibroblasts through multiple signaling pathways involving Rho kinase, ERK, PLC, and phosphoinositide-3-kinase.

Figure 1

Cancer CM-induced secretion of VEGF and SDF-1 from hASCs. (A) hASCs were pretreated with SKOV3 CM, OVCAR3 CM, 5 µM LPA, or serum-free α-MEM for 48 h for collection of SKOV3 CM-hASC CM, OVCAR3 CM-hASC CM, LPA-hASC CM, or hASC CM, respectively. HUVECs were plated onto a layer of Matrigel, and treated with the culture supernatants from hASCs for 12 h. As control experiments, HUVECs were treated with α-MEM (control), SKOV3 CM, OVCAR3 CM, or 5 µM LPA for 12 h. The images were photographed using an inverted microscope with a digital camera. (B and C) hASCs were treated with the indicated concentrations of SKOV3 CM for 48 h for collection of SKOV3 CM-hASC CM. The concentrations of VEGF (B) and SDF-1 (C) in SKOV3 CM-hASC CM or SKOV3 CM were determined by ELISA.

Figure 2

Role of LPA-LPA₁ signaling in the cancer CM-induced expression of α-SMA, VEGF, and SDF-1. (A) Concentrations of LPA in CM from OVCAR3, SKOV3, and hASCs were determined. (B) Serum-starved hASCs were exposed to OVCAR3 CM, SKOV3 CM, or 5 µM 1-oleoyl-LPA in the presence or absence of 1 µM Ki16425 for 4 days. The expression levels of α-SMA and GAPDH were determined by Western blotting. (C-D) hASCs were treated with serum-free medium containing OVCAR3 CM, SKOV3 CM, 5 αM LPA or 10 ng/ml OSM in the presence or absence of 1 αM Ki16425 for 48 h. (E) hASCs were infected with control (sh-control) or LPA₁ (sh-LPA₁) shRNA lentivirus (pLKO.1-puro, MISSION, Sigma). After puromycin selection, the expression levels of LPA₁ and GAPDH were determined by RT-PCR. (F) shRNA-infected cells were exposed to serum-free medium containing 5 µM LPA, OVCAR3 CM or SKOV3 CM for 4 days and the expression levels of α-SMA and GAPDH were determined by Western blotting. (G and H) hASCs were infected with control or LPA1 shRNA lentivirus and treated with serum-free medium containing OVCAR3 CM, SKOV3 CM, 5 µM LPA or 10 ng/ml OSM for 48 h. The secreted levels of VEGF and SDF-1 were determined by ELISA. Data represent mean ± S.D. (n = 4). ^*, P < 0.01 by two-way ANOVA and Scheffe's post hoc test.

Figure 3

Role of LPA₁ in the cancer CM stimulation of hASCs-mediated paracrine function on endothelial tube formation. hASCs were infected with sh-control or sh-LPA₁ lentivirus and then treated with SKOV3 CM, 5 µM LPA, or serum-free α-MEM for 48 h for collection of SKOV3 CM-hASC CM, LPA-hASC CM, or hASC CM, respectively. HUVECs were plated onto a layer of Matrigel, and treated with the conditioned medium from hASCs for 12 h. The images were photographed using an inverted microscope with a digital camera. Representatives of three independent experiments are shown.

Figure 4

Role of myocarin and MRTF-A in the LPA-induced expression of α-SMA, VEGF and SDF-1. (A) hASCs were treated with 5 µM LPA or 1 ng/ml TGF-β1 for the indicated times. (B) hASCs were transfected with control siRNA (si-control), siRNAs specific for myocardin (si-myocardin) or MRTF-A (si-MRTF-A). The mRNA levels of α-SMA, h1-calponin, myocardin, MRTF-A, and GAPDH were determined by RT-PCR. (C) The siRNA-transfected hASCs were treated with 5 µM LPA or vehicles for 4 days and the expression levels of α-SMA and GAPDH were determined by Western blotting. Representatives of three independent experiments are shown. (D and E). The siRNA-transfected hASCs were treated with serum-free media containing 5 µM LPA or 10 ng/ml OSM for 48 h and the CM from hASCs was subjected to ELISA for analysis of VEGF (D) and SDF-1 (E). Data represent mean ± S.D. (n = 4). ^*, P < 0.01 by two-way ANOVA and Scheffe's post hoc test.

Figure 5

Role of RhoA-Rho kinase-dependent pathway in the LPA-induced expression of α-SMA, VEGF and SDF-1. (A) Serum-starved hASCs were treated with 5 µM LPA for the indicated time periods. The amounts of RhoA in the whole cell lysates (total RhoA) or the GTP-bound RhoA precipitated from the lysates were revealed by Western blotting using anti-RhoA antibody. (B) Serum-starved hASCs were treated with 5 µM LPA in the absence or presence of 10 µM Y27632 for 4 days. The expression levels of α-SMA and GAPDH were determined by Western blotting. Representative data from three independent experiments are shown. (C and D) hASCs were treated with serum-free medium containing 5 µM LPA or 10 ng/ml OSM in the absence or presence of 10 µM Y27632 for 48 h and the CM from hASCs was subjected to ELISA for analysis of VEGF (C) and SDF-1 (D). Data represent mean ± S.D. (n = 4). ^*, P < 0.01 by two-way ANOVA and Scheffe's post hoc test.

Figure 6

Signaling pathways involved in the LPA-induced expression of α-SMA, VEGF, and SDF-1. (A) hASCs were treated with 5 µM LPA or vehicles in the absence or presence of 10 µM U0126, 2.5 µM U73122, 10 µM LY294002 for 4 days and the expression levels of α-SMA and GAPDH were determined by Western blotting. (B and C) hASCs were treated with 5 µM LPA or vehicles in the absence or presence of 10 µM U0126, 2.5 µM U73122, 10 µM LY294002 for 2 days and the conditioned medium was subjected to ELISA for analysis of VEGF (B) and SDF-1 (C). Data represent mean ± S.D. (n = 4). ^*, P < 0.01 by two-way ANOVA and Scheffe's post hoc test.