Prostacyclin inhibits non-small cell lung cancer growth by a frizzled 9-dependent pathway that is blocked by secreted frizzled-related protein 1

Abstract

The goal of this study was to assess the ability of iloprost, an orally active prostacyclin analog, to inhibit transformed growth of human non-small cell lung cancer (NSCLC) and to define the mechanism of iloprost's tumor suppressive effects. In a panel of NSCLC cell lines, the ability of iloprost to inhibit transformed cell growth was not correlated with the expression of the cell surface receptor for prostacyclin, but instead was correlated with the presence of Frizzled 9 (Fzd 9) and the activation of peroxisome proliferator-activated receptor-gamma (PPARgamma). Silencing of Fzd 9 blocked PPARgamma activation by iloprost, and expression of Fzd 9 in cells lacking the protein resulted in iloprost's activation of PPARgamma and inhibition of transformed growth. Interestingly, soluble Frizzled-related protein-1, a well-known inhibitor of Wnt/Fzd signaling, also blocked the effects of iloprost and Fzd 9. Moreover, mice treated with iloprost had reduced lung tumors and increased Fzd 9 expression. These studies define a novel paradigm, linking the eicosanoid pathway and Wnt signaling. In addition, these data also suggest that prostacyclin analogs may represent a new class of therapeutic agents in the treatment of NSCLC where the restoration of noncanonical Wnt signaling maybe important for the inhibition of transformed cell growth.

Figure 1

Iloprost inhibits anchorage-independent growth in some but not all NSCLC. (A) The indicated NSCLC cell lines were grown in soft agar in the absence or presence of 10 µM iloprost. The colonies were incubated for several weeks and subsequently counted as described in the Materials and Methods. Data are shown as mean of three independent experiments with the SEM indicated. (B) The cell line A549 is shown to be iloprost-sensitive by soft agar, whereas H460 is seen to be an iloprost-insensitive cell line. (C) Activation of PPARγ and E-cadherin is induced by iloprost. The indicated NSCLC cell lines were transiently transfected with PPAR-RE, along with CMV-β-gal to normalize for transfection efficiency. After an overnight incubation, cells were exposed for 48 hours with 10 µM iloprost. Extracts were prepared and promoter activity was determined as luciferase units normalized to CMV-β-gal. Results represent the mean of three independent experiments with the SEM indicated. (D) The indicated NSCLC cell lines were transiently transfected with E-cadherin reporter, along with CMV-β-gal to normalize for transfection efficiency and exposed for 48 hours with 10 µM iloprost. The PPARγ antagonist compound T0070907 reduces PPAR-RE activity induced by iloprost/Fzd 9. (E) The indicated cell lines were transiently transfected with PPAR-RE, or empty vector, along with CMV-β-gal to normalize for transfection efficiency. After an overnight incubation, cells were exposed for 48 hours with 10 µM iloprost and/or 5 µM T0070907.

Figure 2

RNAi knockdown of Fzd 9 results in reduced PPARγ activity expression. (A) Total RNA purified from the indicated NSCLC cell lines were submitted to quantitative RT-PCR using primers specific for Fzd 9 as described under Materials and Methods. The relative mRNA abundance for Fzd 9 in the different samples was normalized to human GAPDH measured by RT-PCR in the same samples. Aliquots of extracts containing equal protein as measured by the Bradford assay from the indicated cells were resolved by SDS-PAGE and immunoblotted with antibodies to Fzd 9 (100 kDa; Aviva Systems Biology). The filters were stripped and reimmunoblotted for β-catenin (80 kDa; BD Transduction Laboratories) as a loading control. (B) The cell lines H157 and H1703 were transiently transfected with or without Fzd 9 and PPAR-RE along with CMV-β-gal to normalize for transfection efficiency. After an overnight incubation, cells were exposed for 48 hours with 10 µM iloprost. (C) siRNA knockdown of Fzd 9 was performed as described in Materials and Methods on the indicated cell lines. Total RNA purified from these cells were then submitted to quantitative RT-PCR using primers specific for Fzd 9 as previously described. The relative mRNA abundance for Fzd 9 in the different samples was normalized to human GAPDH. (D) The indicated cell lines were transiently transfected with PPAR-RE, along with CMV-β-gal to normalize for transfection efficiency. In addition, Fzd 9 was knocked down by transient siRNA expression as indicated previously. After an overnight incubation, cells were exposed for 48 hours with 10 µM iloprost. (E) The indicated cell line was transiently transfected with PPAR-RE, empty vector, or Fzd 9 along with CMV-β-gal to normalize for transfection efficiency. After an overnight incubation, cells were exposed for 48 hours with 10 µM iloprost and/or 5 µM T0070907.

Figure 3

Fzd 9 is necessary for the effects of iloprost reversal of anchorage independent growth. (A and B) A549 and H661 cells were grown in soft agar exposed to siRNA knockdown of Fzd 9 or the PPARγ antagonist compound T0070907 at a concentration of 5 µM in the absence or presence of 10 µM iloprost. The colonies were incubated for several weeks and subsequently counted as previously described in Materials and Methods. (C and D) The H157 and H1703 cell lines encoding stable empty vector LPCX/LNCX or Fzd 9 was grown in soft agar in the absence or presence of 10 µM iloprost. The colonies were incubated for several weeks and subsequently counted through Metamorph (Downingtown, PA).

Figure 4

Iloprost and Fzd 9 are specific in its activation of PPAR-RE and is not a generalized response to all prostaglandins, the effects are also blocked by sFRP1. (A) Iloprost and Fzd 9 activate PPARγ. The A549 and H1703 cell lines were transiently transfected with the PPAR-RE, sFRP1, Fzd 9, Wnt 7a, along with CMV-β-gal to normalize for transfection efficiency. After an overnight incubation, cells were exposed for 48 hours with 10 µM iloprost. (B) The H157 cell line was transiently transfected with PPAR-RE, Fzds (1–10), along with CMV-β-gal to normalize for transfection efficiency. After an overnight incubation, cells were exposed for 48 hours with 10 µM iloprost. (C) The cell line H157 was transiently transfected with the reporter plasmid PPAR-RE and the effector plasmid PGIS, or Fzd 9, or exposed to 10 µM PGE₂. Of note, CMV-β-gal was used to normalize for transfection efficiency. The cells were incubated for 48 hours, and luciferase and β-galactosidase activities were measured. Data are presented as relative light units/milliunit of β-galactosidase activity. The results showed that the coexpression of PGIS and Fzd 9 stimulated PPAR-RE activity but not the sole expression of PGIS, Fzd 9, or PGE₂. Furthermore, the coexpression of PGE₂ and Fzd 9 similarly did not stimulate PPAR-RE activity.

Figure 5

Iloprost and Fzd 9 induce phospho-ERK5 and MEF2C activity in NSCLC and is associated with PPARγ activity. (A) Retroviruses encoding stable empty LNCX/LPCX or Fzd 9 were used to transduce the H157 NSCLC cell line as described previously in Materials and Methods. Extracts were prepared from pooled G418 and puromycin-resistant cultures with MAPK lysis buffer, and aliquots containing 100 µg of protein were resolved on 10% polyacrylamide SDS gels, transferred to nitrocellulose, and probed with an antibody to phospho-ERK5 (115 kDa; Cell Signaling). The filters were stripped and reimmunoblotted for total ERK5 (115 kDa; Cell Signaling), which was used as a loading control protein. (B) The H157 cell line was transiently transfected with the MEF2C reporter, along with CMV-β-gal to normalize for transfection efficiency. MKK5 alpha DD was used as a positive control plasmid. After an overnight incubation, cells were exposed for 48 hours with 10 µM iloprost. (C) Retroviruses encoding empty LNCX were used to transduce the A549 NSCLC cell line as described previously in Materials and Methods. Extracts were prepared from pooled G418-resistant cultures with MAPK lysis buffer and aliquots containing 100 µg of protein were resolved on 10% polyacrylamide SDS gels, transferred to nitrocellulose, and probed with an antibody to phospho-ERK5 (115 kDa; Cell Signaling). The filters were stripped and reimmunoblotted for total ERK5 (115 kDa; Cell Signaling), which was used as a loading control protein. (D) The A549 cell line was transiently transfected with the MEF2C reporter, along with CMV-β-gal to normalize for transfection efficiency. MKK5 alpha DD was used as a positive control plasmid. After an overnight incubation, cells were exposed for 48 hours with 10 µM iloprost.

Figure 6

Iloprost treatment reduces tumor number and increases whole lung expression of Fzd 9. (A) Representative images of hematoxylin and eosin-stained urethane-treated lung tissue showing reduced tumor numbers and size in lungs from iloprost-fed mice (bottom) compared with control chow-fed mice (top). The thick arrows indicate tumor(s) in the lung and the thin arrow points out the more normal lung architecture in the mice receiving 3% iloprost chow. The right panels are higher-magnification images from the insets on the left, demonstrating the effect of iloprost on individual adenoma size. Magnifications: x100 (left) and x400 (right). (B) Whole lung tissue from urethane-treated mice fed normal or iloprost-impregnated chow (n = 4 each group) was snap-frozen, and whole-cell lysates were analyzed by Western blot analysis for Fzd 9 expression. β-Actin was used as a loading control. (C) Lungs from mice described in panel A were removed and tumors were counted. *P = .0002, different from control chow-fed mice.