LASS2 impairs proliferation of glioma stem cells and migration and invasion of glioma cells mainly via inhibition of EMT and apoptosis promotion

Abstract

LAG1 longevity assurance homolog 2 (LASS2), a highly conserved transmembrane protein, has been reported in several cancer types. However, the roles of LASS2 in glioma biology remain elusive. In the present study, we investigated the expression of LAAS2 in human glioma tissues and the effects of LASS2 on glioma stem cell (GSC) proliferation. Roles of LASS2 in glioma cell migration and invasion were also researched both in vitro and in vivo. Our results demonstrated that the level of LASS2 is gradually reduced with the increase of glioma grade. The level of LASS2 is significantly lower in GSCs than in non GSCs, whereas LASS2 overexpression reduced the sphere formation and promoted the differentiation of CD133⁺ glioblastoma cells, as was indicated by reduced levels of CD133 and Nestin. In addition, LASS2 overexpression significantly reduced colony formation, migration, and invasion of glioma cells by promoting tumor cell apoptosis and inhibiting epithelial-mesenchymal transition (EMT). Overexpression of LASS2 inhibited U-87 MG cell-derived glioma xenograft growth in nude mice in a manner similar to in vitro. Our findings indicate that LASS2 can function as a suppressor of glioma growth, suggesting that modulation of LASS2 expression may contribute to a novel strategy for the management of glioma via inhibition of GSCs.

Keywords: LASS2; epithelial-mesenchymal transition (EMT); glioblastoma; glioma; glioma stem cells; invasion; migration.

Figure 1

Analysis of LASS2 expression on a human glioma tissue microarry. (A) Representative immunohistochemical staining images of LASS2. Scale bar = 100 µm. (B) LASS2 levels in grade II to IV glioma samples were significantly lower than those in the normal adjancent tissue (NAT) (*P < 0.05 for grade II, and **P < 0.01 for both grade III and IV, n.s., no significance; one-way ANOVA). (C) The level of LASS2 in glioma of all grades was significantly lower than in NAT (*P < 0.05; unpaired two-tailed Student's t-test). (D) Representative images showing the immunuofluorescent co-staining of LASS2 and CD133 in the NAT and gliomas graded from I to IV. Scale bar = 20 µm.

Figure 2

Isolation of GSCs from U-87 MG and U251 cells. (A) The proportion of CD133⁺ cells in U-87 MG and U251 cells. (B) The sorted CD133⁺ U-87 MG cells were retested by flow cytometry after 7 days of culture, and the percentages of CD133⁺ cells and CD133^- cells were shown. (C) CD133⁺ cells and CD133^- cells were maintained in the serum-free condition for 7 days. CD133⁺ cells formed spheres, whereas CD133^- cells showed degradation. Scale bar = 200 µm. (D) The sphere derived from sorted CD133⁺ cells was shown to express stem cell markers CD133 and Nestin. Scale bar = 20 µm. (E) CD133 and Nestin were overexpressed in proliferated sorted CD133⁺ cells compared with CD133^- cells, whereas CD133^- cells overexpressed LASS2 compared with CD133⁺ cells (*P < 0.05, vs. CD133^- group; unpaired two-tailed Student's t-test; n = 3). (F, G) The glioma xenografts derived from CD133⁺ cells were more malignant than those derived from CD133- cells. (H) The average final tumor weight of the xenograft derived from CD133⁺ cells was significantly higher than that from CD133^- cells in nude mice (F, G, and H, *P < 0.05, **P < 0.01, and ***P < 0.001, vs. CD133^- group; unpaired two-tailed Student's t-test; n = 3 animals).

Figure 3

LASS2 inhibits sphere formation of GSCs. CD133⁺ U-87 MG cells were stably transfected with pLV and pLV-LASS2. (A) Downregulation of LASS2 promoted sphere formation of GSCs (**P < 0.01, vs. pLV control; unpaired two-tailed Student's t-test; n = 3). Scale bar = 200 µm. (B) Downregulation of LASS2 significantly reduced the protein level of LAAS2 while significantly increasing those of Notch1 and glioma stem cells stemness proteins CD133, Nestin and Sox2 (*P < 0.05, vs. pLV control; unpaired two-tailed Student's t-test; n = 3).

Figure 4

Effect of LASS2 on cell migration, invasion and apoptosis. (A) Wound healing assay of pLV-vector or pLV-LASS2-transfected glioma/glioblastoma cells at 0 h, 12 h and 24 h after scratch. The images were taken from an inverted microscope under 10× magnification (*P < 0.05 and **P < 0.01, vs. pLV control; unpaired two-tailed Student's t-test; n = 3). Scale bar = 200 µm. (B) Colony formation assay in pLV-vector or pLV-LASS2-transfected U251 and U-87 MG cells. Images were acquired at 4× magnification (*P <0.05, vs. pLV control; unpaired two-tailed Student's t-test; n = 3). (C) Transwell assay demonstrated that LASS2 inhibits the migration of U251 and U-87 MG cells compared with the pLV control (*P <0.05; unpaired two-tailed Student's t-test; n = 3). Scale bar = 200 µm. (D) The immunofluorescence staining of MMP9 and SPHK1 was shown in both U251 and U-87 MG cells. Scale bar = 200 µm. (E) RNA-Seq shows that LASS2 influenced cell migration/invasion, apoptosis, epithelial- mesenchymal transition (EMT) conversion and cellular life activity. (F) Overexpression of LASS2 reduced the protein levels of MMP2, MMP9, and SPHK1 while increasing that of TIMP2 in both U251 and U-87 MG cells. (G) LASS2 overexpression increased the levels of Bax, cleaved Caspase-3, TNF-α, and p53 while reducing that of Bcl-2 in both U251 and U-87 MG cells. (H) Overexpression of LASS2 reduced the protein levels of Vimentin and N-cadherin while increasing that of E-cadherin in both U251 and U-87 MG cells (F, G, and H, *P <0.05 and **P < 0.01, vs. pLV control in each cell line; unpaired two-tailed Student's t-test; n = 3).

Figure 5

LASS2 inhibited tumor growth in a pLV-LASS2-U-87 MG glioblastoma xenograft nude mouse model. (A) Representative photographs showing the gross pLV-LASS2-U-87 MG and empty scramble control glioblastoma xenografts from the nude mouse. (B) The tumor volume was evaluated between the scrambled control and pLV-LASS2 groups (**P < 0.05 and **P < 0.01 vs. pLV control group; unpaired two-tailed Student's t-test; n = 5 animals). (C) The final tumor weight was measured after dissection. The average final weight of tumors derived from pLV-LASS2-transcfected U-87 MG cells was significantly lower than those derived from the scrambled control (**P < 0.01, vs. pLV control group; unpaired two-tailed Student's t-test; n = 5 animals). (D) Representative images for H&E staining from either group were shown. (E) IHC staining of LASS2, TIMP2, MMP9, and SPHK1 in xenografted tumors derived from U-87 MG cells transfected with either pLV-LASS2 or scrambled control. Scale bar = 20 µm. (F) Western blot analysis of LASS2, SPHK1, TIMP2, MMP2, and MMP9 in xenografted tumors derived from U-87 MG cells transfected with either pLV or pLV-LASS2. (G) Western blot analysis of Bax, Bcl-2, pro-Caspase-3, cleaved Caspase-3, TNF-α and p53 in xenografted tumors derived from U-87 MG cells transfected with either pLV or pLV-LASS2. (H) Western blot analysis of EMT conversion-related proteins Vimentin, E-cadherin, and N-cadherin in xenografted tumors derived from U-87 MG cells transfected with either pLV or pLV-LASS2 (F, G, and H, *P <0.05, **P < 0.01, and ***P < 0.001, vs. pLV control; unpaired two-tailed Student's t-test; n = 3).