Cav-1 Ablation in Pancreatic Stellate Cells Promotes Pancreatic Cancer Growth through Nrf2-Induced shh Signaling

Abstract

A more comprehensive understanding of the complexity of pancreatic cancer pathobiology, especially, and understanding of the role of the tumor microenvironment (TME) in disease progression should pave the way for therapies to improve patient response rates. Previous studies reported that caveolin-1 (Cav-1) has both tumor-promoting and tumor-suppressive functions. However, the function of Cav-1 in the pancreatic cancer microenvironment remains largely unexplored. Here, we show that coinjection of Cav-1-silenced pancreatic stellate cells (PSCs) with pancreatic cancer cells increased tumor growth. To comprehensively characterize paracrine communication between pancreatic cancer cells and PSCs, PSCs were cultured with pancreatic cancer cell conditioned medium (CM) containing cytokines. We reveal that Cav-1-silenced PSCs facilitated the growth of pancreatic cancer cells via enhanced paracrine shh/MMP2/bFGF/IL-6 signaling. Specifically, Cav-1-silenced PSCs exhibited increased shh expression, which heterotypically activated the shh signaling pathway in pancreatic cancer cells. Moreover, Cav-1-deficient PSCs accumulated ROS to enhance the shh pathway and angiogenesis in pancreatic cancer cells. In addition, overexpression of Nrf2 reversed the effects of Cav-1 knockdown on PSCs, increasing ROS production and enhancing paracrine shh/MMP2/bFGF/IL-6 signaling. Together, our findings show that stromal Cav-1 may mediate different mechanisms in the complex interaction between cancer cells and their microenvironment though Nrf2-induced shh signaling activation during pancreatic cancer progression.

Figure 1

Cav-1 interference in PSCs facilitates Aspc-1 pancreatic cancer cell growth. (a) Cav-1 interference efficiency in PSCs was analyzed by qRT-PCR (left). Cav-1 and α-SMA protein expression in Cav-1-knockdown PSCs was analyzed by western blot (middle and right). sh-Ctrl stands for control shRNA. sh-Cav-1 stands for Cav-1 shRNA. (b) A total of 5 × 10⁶ Aspc-1 cells mixed with 0.5 × 10⁶ PSCs with or without Cav-1 knockdown (sh-Ctrl or sh-Cav-1) were implanted subcutaneously into the flanks of BALB/c nude mice (n = 6 per group). Tumor volumes were determined by measuring the width and length of the tumors every week. Mean (n = 6); bars, SD; representative images of tumors are displayed in (c). (d and e) CD31 immunoblot analysis of whole-tumor lysates is shown. (f and g) CD31 immunohistochemistry of tumor sections showing that microvascular density correlates with tumor size in the Cav-1-knockdown PSC group. The results shown are the means ± SEMs. ^∗P < 0.05, by the two-tailed Mann–Whitney test and by Dunnett's multiple comparison test.

Figure 2

Cav-1-knockdown PSCs exhibit increased amounts of protumorigenic cytokines that promote pancreatic cancer growth and invasion. (a) shh, bFGF, MMP2, and IL-6 secretion in the CM of PSCs with or without Cav-1 knockdown (sh-Ctrl or sh-Cav-1) was determined by ELISA. (b and c) The results of shh immunoblot analysis of PSCs with or without Cav-1 knockdown (sh-Ctrl or sh-Cav-1) is displayed. (d) Aspc-1 cells were incubated with CM from PSCs with or without Cav-1 knockdown (sh-Ctrl or sh-Cav-1). The invasive ability of Aspc-1 cells was evaluated by counting the numbers of migrated cells in ten randomly selected fields under a light microscope at ×100 magnification. (e and f) MTT assay and [³H]thymidine incorporation assay (48 hours) in Aspc-1 pancreatic cancer cells treated with CM from PSCs with or without Cav-1 knockdown. (g) Immunoblot analysis showing increased expression of Gli-1, cyclin D1, and cyclin A in Aspc-1 pancreatic cancer cells incubated (48 hours) with CM from Cav-1-knockdown PSCs. Data represent the means ± SEMs. ^∗P < 0.05, by two-tailed Student's t-test.

Figure 3

Suppression of shh signaling in pancreatic cancer cells abrogates the proproliferative/protumorigenic effects of Cav-1-knockdown PSCs. (a) [³H]Thymidine incorporation assay of Aspc-1 pancreatic cancer cells treated with DMSO or cyclopamine after incubation with CM from PSCs (sh-Ctrl or sh-Cav-1). ^∗P < 0.05 versus the sh-Ctrl group; ^#P < 0.05 versus the sh-Cav-1 group treated with DMSO. (b and c) Gli-1 and β-actin immunoblot analysis of sh-Ctrl- or sh-Gli-1-transfected Aspc-1 cells before and after treatment with shh. (d) A total of 5 × 10⁶ Aspc-1-sh-Gli-1 cells mixed with 0.5 × 10⁶ PSCs with or without Cav-1 knockdown (sh-Ctrl or sh-Cav-1) were implanted subcutaneously in the flanks of BALB/c nude mice (n = 6 per group). sh-Ctrl stands for coinjection of Aspc-1 cells and sh-Ctrl-transfected PSC group; sh-Gli-1 stands for coinjection of sh-Gli-1-transfected Aspc-1 cells and sh-Ctrl-transfected PSC group; sh-Cav-1 stands for coinjection of Aspc-1 cells and sh-Cav-1-transfected PSC group; sh-Gli-1+sh-Cav-1 stands for coinjection of sh-Gli-1-transfected Aspc-1 cells and sh-Cav-1-transfected PSC group; tumor volumes were determined by measuring the width and length of the tumors every week. Mean (n = 6); bars, SD; Gli-1 knockdown in Aspc-1 cells reversed the tumor-promoting effects of Cav-1-knockdown PSCs, as determined by coinjection experiments. The results are the means ± SEMs. n = 6 per group; ^∗P < 0.05, by Tukey's multiple comparison test.

Figure 4

ROS are a key mediator in the observed effects of Cav-1-knockdown PSCs on shh signaling and angiogenesis in pancreatic cancer. (a) ROS production in PSCs with or without Cav-1 knockdown was evaluated with DCF-DA and normalized based on total protein content. (b) Gli-1 and β-actin immunoblot analysis of sh-Ctrl- or sh-Gli-1-transfected Aspc-1 cells cultured with CM from sh-Ctrl- or sh-Cav-1-transfected PSCs before and after treatment with NAC. (c and d) Angiogenesis was evaluated based on tube formation. HUVECs were cocultured with CM from the indicated groups. sh-Ctrl stands for CM from Aspc-1 cells cocultured with sh-Ctrl-transfected PSCs; sh-Ctrl+NAC stands for CM from Aspc-1 cells cocultured with sh-Ctrl-transfected PSCs and 20 mM NAC; sh-Gli-1 stands for CM from sh-Gli-1-transfected Aspc-1 cells cocultured with sh-Ctrl-transfected PSCs; sh-Gli-1+NAC stands for CM from sh-Gli-1-transfected Aspc-1 cells cocultured with sh-Ctrl-transfected PSCs and 20 mM NAC; sh-Cav-1 stands for CM from Aspc-1 cells cocultured with sh-Cav-1-transfected PSCs; sh-Cav-1+NAC stands for CM from Aspc-1 cells cocultured with sh-Cav-1-transfected PSCs and 20 mM NAC; sh-Cav-1+sh-Gli-1 stands for CM from sh-Gli-1-transfected Aspc-1 cells cocultured with sh-Cav-1-transfected PSCs; tube numbers were counted. Results are shown as the means ± SEMs; ^∗P < 0.05, by two-tailed Student's t-test.

Figure 5

Nrf2 mediates ROS production induced by Cav-1 knockdown in PSCs. (a) Nrf2 protein levels in PSCs were determined by western blot. (b) GSH and GSSG levels were evaluated in PSCs (middle and right). ^∗P < 0.05 versus the sh-Ctrl group (n = 3). All data are representative of at least three independent experiments. (c–e) Cav-1 in PSCs was silenced by transfection with control shRNA (sh-ctrl) or shRNA targeting Cav-1 (sh-Cav-1). Nrf2 overexpression DNA (Nrf2 vector) or control DNA (control vector) was added to the Cav-1-silenced group to determine the effect of Nrf2 overexpression on Cav-1-knockdown PSCs. Nrf2 and shh protein expression was detected by western blot. (f) ROS production in PSCs was evaluated with DCF-DA and normalized based on total protein content. Data represent the means ± SEMs. ∗ denotes P < 0.05, by two-tailed Student's t-test.

Figure 6

Nrf2 regulates shh/MMP-2/bFGF/IL-6 secretion elevated by Cav-1 knockdown in PSCs. Cav-1 in PSCs was silenced by transfection with control shRNA (sh-ctrl) or shRNA targeting Cav-1 (sh-Cav-1). Nrf2 overexpression DNA (Nrf2 vector) or control DNA (control vector) was added to the Cav-1-silenced group to determine the effect of Nrf2 overexpression on Cav-1 knockdown PSCs. The secretion of shh (a), MMP2 (b), bFGF (c), and IL-6 (d) was determined by ELISA. The results are shown as the means ± SEMs. ∗ denotes P < 0.05, by two-tailed Student's t-test.