LRG1 Suppresses Migration and Invasion of Esophageal Squamous Cell Carcinoma by Modulating Epithelial to Mesenchymal Transition

Abstract

Background: Esophageal squamous cell carcinoma (ESCC) is a common cancer with poor prognosis. The molecular pathogenesis underlying ESCC remains to be explored. Leucine-rich ɑ-2-glycoprotein 1 (LRG1) has been implicated in the pathogenesis of various cancer types, however its role in ESCC is unknown. Materials and Methods: Data from the public database was analyzed to address the expression of LRG1 in ESCC. Gain-of-function studies were performed in select ESCC cell lines by over-expression or addition of recombinant LRG1, while loss-of-function studies achieved by small interfering RNA mediated knockdown. Wound healing and transwell assays were conducted to investigate ESCC cell migration and invasion upon manipulating LRG1 levels. Western blot and Immunofluorescence staining were used to examine the changes in epithelial to mesenchymal transition (EMT) and TGFβ signaling pathway. Results: LRG1 mRNA levels were found to be significantly down-regulated in patients with ESCC as well as in several ESCC cell lines. Silencing of LRG1 promoted, while overexpression of LRG1 inhibited ESCC cell migration and invasion. In line with this, Silencing of LRG1 enhanced, while overexpression of LRG1 reduced TGFβ signaling and EMT of ESCC cells. Conclusion/Significance: LRG1 suppresses ESCC cell migration and invasion via negative modulation of TGFβ signaling and EMT. Down-regulation of LRG1 in ESCC patients may favor tumor metastasis and disease progression.

Keywords: Epithelial to Mesenchymal Transition; Esophageal Squamous Cell Carcinoma; Invasion; LRG1; Migration; TGFβ signaling.

Fig 1

Expression levels of LRG1 in patients with ESCC and select ESCC cell lines. (A) Comparison of the mRNA levels of LRG1 in tumor and adjacent normal tissue samples from 51 patients with ESCC. The analysis was done with dataset GSE23400 from the GEO database, and the mRNA levels were reflected by the median centered intensity of the original profiling signal. *** indicates P < 0.001. (B) The mRNA levels of LRG1 in select ESCC cell lines as revealed by real-time PCR. Three randomly chosen ESCC cell lines (TE1, KYSE30 and EC109) were tested along with the normal esophageal tissue cell HET-1A and the hepatoma line HepG2 as positive control. The relative mRNA levels of LRG1 in HET-1A was set as 1. (C) The protein levels of LRG1 in the select ESCC cell lines as revealed by Western blot. β-actin was used as loading control.

Fig 2

Effect of LRG1 knockdown on ESCC cell migration and invasion. (A) Western blot analysis of LRG1 protein levels in KYSE30 cells after transfection of 3 different siRNAs of LRG1 or a scramble negative control (NC). β-actin was used as loading control. NC or siLRG1 #2 transfected cells were starved for 12 h before wound-healing or transwell assay. (B) For wound-healing, cells were manually scratched using a pipette tip, washed and maintained in serum-free culture medium. Pictures were taken at 0 h and 24 h after scratching (left), and the wound closure ratio (cell migration distance at 24 h divided by the gap distance at 0 h) was obtained for comparison (right). Data were representative of three independent experiments and shown as mean ± SD. ** indicates P < 0.01. (C) For transwell assay, cells were added into the upper chamber in serum-free culture medium. The chamber was then placed in a well of 24-well plate that was filled with FBS-containing complete medium. 36 hours later, migrated cells were fixed and stained with DAPI for imaging. The invasion assay was performed similarly except that matrigel coated chambers was used (left). The number of migrated or invaded cells between the two groups were compared (right). Data were representative of three independent experiments and shown as mean ± SD.* indicates P < 0.05, ** indicates P < 0.01.

Fig 3

Effect of LRG1 overexpression on ESCC cell migration and invasion. (A) Western blot analysis of LRG1 protein levels in EC109 cells after transfection of LRG1-overexpressing plasmid (pCMV-LRG1) or the empty vector. β-actin was used as loading control. Vector or pCMV-LRG1 transfected cells were starved in serum free medium for 12 h before wound-healing or transwell assay. (B) For wound-healing, cells were manually scratched using a pipette tip, washed and maintained in serum-free culture medium. Pictures were taken at 0 h and 24 h after scratching (left), and the wound closure ratio (cell migration distance at 24 h divided by the gap distance at 0 h) was obtained for comparison (right). Data were representative of three independent experiments and shown as mean ± SD.** indicates P < 0.01. (C) For transwell assay, cells were added into the upper chamber in serum-free culture medium. The chamber was then placed in a well of 24-well plate that was filled with FBS-containing complete medium. 36 hours later, migrated cells were fixed and stained with DAPI for imaging. The invasion assay was performed similarly except that matrigel coated chambers was used (left). The number of migrated or invaded cells between the two groups were compared (right). Data were representative of three independent experiments and shown as mean ± SD.*** indicates P < 0.001.

Fig 4

Effect of recombinant LRG1 on ESCC cell migration and invasion. EC109 cells were treated with recombinant human LRG1 at indicated concentration and starved in serum free medium for 12 h before wound-healing or transwell assay. (A) For wound-healing, cells were manually scratched using a pipette tip, washed and maintained in serum-free culture medium. Pictures were taken at 0 h and 24 h after scratching. (B) For transwell assay, cells were added into the upper chamber in serum-free culture medium. The chamber was then placed in a well of 24-well plate that was filled with FBS-containing complete medium. 36 hours later, migrated cells were fixed and stained with DAPI for imaging. The invasion assay was performed similarly except that matrigel coated chambers was used. Data were representative of three independent experiments.

Fig 5

Effect of LRG1 knockdown on the epithelial to mesenchymal transition (EMT) of ESCC cells. (A) Western blot analysis of protein expression of EMT marker genes upon LRG1 knockdown in KYSE30 cells with siRNA #2 and #3. β-actin was used as loading control. (B) Immunofluorescence staining of E-cadherin (top) and N-cadherin (bottom) in NC or siLRG1 #2 transfected cells. Data were representative of three independent experiments.

Fig 6

Effect of LRG1 overexpression on the epithelial to mesenchymal transition (EMT) of ESCC cells. (A) Western blot analysis of protein expression of EMT marker genes upon LRG1 overexpression in EC109 cells. β-actin was used as loading control. (B) Immunofluorescence staining of E-cadherin (top) and N-cadherin (bottom) in vector or pCMV-LRG1 transfected cells. Data were representative of three independent experiments.

Fig 7

Manipulation of LRG1 levels modulated the activation of TGFβ pathway. Shown were western blot analysis of protein expression of TGF-β1, SMAD2/3 and phosphorylated SMAD2/3 upon LRG1 knockdown in KYSE30 cells with siRNA #2 and #3 (left), or upon LRG1 overexpression in EC109 cells (right). β-actin was used as loading control. Data were representative of three independent experiments.