A novel molecular pathway for Snail-dependent, SPARC-mediated invasion in non-small cell lung cancer pathogenesis

Abstract

Definition of the molecular pathogenesis of lung cancer allows investigators an enhanced understanding of the natural history of the disease, thus fostering development of new prevention strategies. In addition to regulating epithelial-to-mesenchymal transition (EMT), the transcription factor Snail exerts global effects on gene expression. Our recent studies reveal that Snail is upregulated in non-small cell lung cancer (NSCLC), is associated with poor prognosis, and promotes tumor progression in vivo. Herein, we demonstrate that overexpression of Snail leads to the upregulation of secreted protein, acidic and rich in cysteine (SPARC) in models of premalignancy and established disease, as well as in lung carcinoma tissues in situ. Snail overexpression leads to increased SPARC-dependent invasion in vitro, indicating that SPARC may play a role in lung cancer progression. Bioinformatic analysis implicates transforming growth factor beta (TGF-β), extracellular signal-regulated kinase (ERK)1/2, and miR-29b as potential intermediaries in Snail-mediated upregulation of SPARC. Both the TGF-β1 ligand and TGF-β receptor 2 (TGF-βR2) are upregulated following Snail overexpression. Treatment of human bronchial epithelial cell (HBEC) lines with TGF-β1 and inhibition of TGF-β1 mRNA expression modulates SPARC expression. Inhibition of MAP-ERK kinase (MEK) phosphorylation downregulates SPARC. MiR-29b is downregulated in Snail-overexpressing cell lines, whereas overexpression of miR-29b inhibits SPARC expression. In addition, miR-29b is upregulated following ERK inhibition, suggesting a Snail-dependent pathway by which Snail activation of TGF-β and ERK signaling results in downregulation of miR-29b and subsequent upregulation of SPARC. Our discovery of pathways responsible for Snail-induced SPARC expression contributes to the definition of NSCLC pathogenesis.

Figure 1. Snail overexpression is correlated with upregulation of SPARC

(A) The NSCLC cell lines A549, H1437, and H292 were stably transfected with either a Vector control plasmid (V) or a Snail expression plasmid (S). Total RNA was isolated from the cell lines and expression levels of SPARC were evaluated by q-RT-PCR using TaqMan primers. mRNA levels were normalized to GUSB. (B) The same NSCLC cell lines were evaluated for protein level expression of Snail and SPARC by western blotting. Protein levels were normalized to α-Tubulin. (C) Total RNA was isolated from HBEC3-V/S, HBEC4-V/S, and H3mut-V/S cell lines. Expression levels of SPARC were evaluated by q-RT-PCR using TaqMan primers. mRNA levels were normalized to GUSB. A similar pattern was also observed for HBEC7 (data not shown). (D) The same HBEC cells were evaluated for protein level expression of Snail and SPARC by western blotting. Protein levels were normalized to α-Tubulin. (E) Serial sections of ADC (a–d). Positive (brown staining) cytoplasmic expression of SPARC (row 1; a and b) and positive nuclear expression of Snail (row 2; c and d) were observed in the epithelial component of the neoplasm; × 100 low magnification (left; a and c) and × 200 high magnification (right; b and d) images of the same specimen and staining condition. Serial sections of SCC (e–h). Positive cytoplasmic expression of SPARC (row 3; e and f, arrows) and positive nuclear expression of Snail (row 4; g and h) were observed in the epithelial component of the neoplasm; × 100 low magnification (left; e and g) and × 400 high magnification (right; f and h) images of the same specimen and staining condition. Necrotic areas of the neoplasm (N) do not stain for either protein. (** = p < 0.001; *** = p < 0.0001)

Figure 2. Snail overexpression leads to SPARC-dependent increased invasion in premalignant and established NSCLC

(A) The invasive capacity of the HBEC lines HBEC3, HBEC4, and H3mutP53/KRAS (H3mut) with and without Snail overexpression were evaluated in a modified Boyden chamber assay for invasion through a collagen matrix over 48 hrs. Fluorescence values were divided by maximum input fluorescence measured on Day 0 for each cell line to derive a percent input invasion value. (B) The NSCLC cell lines A549, H1437, and H292 with and without Snail overexpression were evaluated as in (A). (C–F) SPARC shRNA sequences (sh1, sh2, sh3) were stably transfected into H3mut and H1437 vector control and Snail-overexpressing cell lines along with a nonsilencing (NS) shRNA control. Protein level expression of SPARC, Snail, and α-Tubulin were evaluated in the HBEC (C) and NSCLC (D) lines by western blot. The HBEC (E) and NSCLC lines (F) were also evaluated in a modified Boyden chamber assay for invasion through a collagen matrix over 48 hrs. The following comparisons were made for both cell types: (1) Vector versus Snail cells transduced with shV and (2) Snail cells transduced with shNS versus shSPARC2. (* = p < 0.05; ** = p < 0.001; *** = p < 0.0001)

Figure 3. TGF-β1 is upregulated by Snail upstream of ERK1/2 and SPARC

(A) The secreted protein levels of TGF-β1 were measured by ELISA from supernatants of Snail overexpressing HBEC lines and compared to appropriate vector controls. (B) Parental HBEC lines were treated with recombinant TGF-β1 (5ng/mL) or vehicle control for 24 hours in serum-free media. Lysates were collected and protein expression of Snail, phosphorylated ERK1/2 (pERK), total ERK1/2, SPARC, and α-tubulin were measured by western blot. Protein levels were normalized to α-tubulin. (C) Snail-overexpressing HBEC lines and vector controls were treated with single siRNA sequences targeting TGF-β1 (si1 and si2), a negative control siRNA (N), or left untreated (V) for 24 hours in serum-free media. Lysates were collected and protein expression was measured as in (B). (D) Efficiency of TGF-β1 knockdown was measured by ELISA 24 hours following transfection. (*** = p < 0.0001)

Figure 4. ERK1/2 is phosphorylated downstream of Snail and upstream of SPARC

(A) The cell lines HBEC2, HBEC3, HBEC7, and H3mutP53/KRAS (H3mut) with and without Snail overexpression (-V/-S) were treated with the MEK1/2 phosphorylation inhibitor U0126 (15uM) and evaluated for SPARC protein expression. Membranes were incubated with antibodies against Snail, pERK, total ERK1/2, SPARC, and α-tubulin. Protein levels were normalized to α-tubulin. (B) SPARC mRNA expression was evaluated following U0126 treatment as in (A). mRNA expression was normalized to GUSB. (C) miR-29b miRNA expression was evaluated following U0126 treatment as in (A). miRNA expression was normalized to RNU6b. (** = p < 0.001; *** = p < 0.0001)

Figure 5. miR-29b is downregulated in NSCLC cell lines overexpressing Snail

(A) Total RNA was isolated from HBEC3-V/S, HBEC4-V/S, and H3mut-V/S cell lines. Expression levels of miR-29b were evaluated by q-RT-PCR using TaqMan primers. miRNA levels were normalized to RNU6b. (B) Total RNA was isolated from A549-V/S, H1437-V/S, and H292-V/S cell lines. Expression levels of miR-29b were evaluated by q-RT-PCR using TaqMan primers. miRNA levels were normalized to RNU6b. (C) The HBEC lines in (A) were stably transfected with a miR-29b precursor sequence and evaluated for expression of Snail and SPARC. Protein levels were normalized to α-tubulin. (D) We propose a regulatory pathway wherein Snail upregulates TGF-β in an autocrine or paracrine fashion, leading to activation of the MEK/ERK pathway, downregulation of miR-29b, and finally upregulation of SPARC. Snail may suppress miR-29b in both an ERK-dependent or -independent manner. (** = p < 0.001; *** = p < 0.0001)