BACKGROUNDRecent studies have suggested that cancer stem cells are one of the major causes for tumor recurrence due to their resistance to radiotherapy and chemotherapy. Although the highly invasive nature of glioblastoma (GBM) cells is also implicated in the failure of current therapies, it is not clear how glioma stem cells (GSCs) are involved in invasiveness. Rac1 activity is necessary for inducing reorganization of actin cytoskeleton and cell movement. In this study, we aimed to investigate the distribution characteristics of CD133+ cells and Rac1+ cells in GBM as well as Rac1 activity in CD133+ GBM cells, and analyze the migration and invasion potential of these cells.

METHODSA series of 21 patients with GBM were admitted consecutively and received tumor resection in Tianjin Medical University General Hospital during the first half of the year 2011. Tissue specimens were collected both from the peripheral and the central parts for each tumor under magnetic resonance imaging (MRI) navigation guidance. Immunohistochemical staining was used to detect the CD133+ cells and Rac1+ cells distribution in GBM specimens. Double-labeling immunofluorescence was further used to analyze CD133 and Rac1 co-expression and the relationship between CD133+ cells distribution and Rac1 expression. Serum-free medium culture and magnetic sorting were used to isolate CD133+ cells from U87 cell line. Rac1 activation assay was conducted to assess the activation of Rac1 in CD133+ and CD133 - U87 cells. The migration and invasive ability of CD133+ and CD133 - U87 cells were determined by cell migration and invasion assays in vitro. Student's t-test and one-way analysis of variance (ANOVA) test were used to determine statistical significance in this study.

RESULTSIn the central parts of GBMs, CD133+ cells were found to cluster around necrosis and occasionally cluster around the vessels under the microscope by immunohistological staining. In the peripheral parts of the tumors, CD133+ cells were lined up along the basement membrane of the vessels and myelinated nerve fibers. Rac1 expression was high and diffused in the central parts of the GBMs, and the Rac1+ cells were distributed basically in accordance with CD133+ cells both in the central and peripheral parts of GBMs. In double-labeling immunofluorescence, Rac1 was expressed in (83.14 ± 4.23)% of CD133+ cells, and CD133 and Rac1 co-expressed cells were located around the vessels in GBMs. Significantly higher amounts of Rac1-GTP were expressed in the CD133+ cells (0.378 ± 0.007), compared to CD133- cells (0.195 ± 0.004) (t = 27.81; P < 0.05). CD133+ cells had stronger ability to migrate (74.34 ± 2.40 vs. 38.72 ± 2.60, t = 42.71, P < 0.005) and invade (52.00 ± 2.28 vs. 31.26 ± 1.82, t = 30.76, P < 0.005), compared to their counterpart CD133- cells in transwell cell migration/invasion assay.

CONCLUSIONSThese data suggest that CD133+ GBM cells highly express Rac1 and have greater potential to migrate and invade through activated Rac1-GTP. The accordance of distribution between Rac1+ cells and CD133+ cells in GBMs implies that Rac1 might be an inhibited target to prevent invasion and migration and to avoid malignant glioma recurrence.