Glioma stem cells are more aggressive in recurrent tumors with malignant progression than in the primary tumor, and both can be maintained long-term in vitro

Background: Despite the advances made during decades of research, the mechanisms by which glioma is initiated and established remain elusive. The discovery of glioma stem cells (GSCs) may help to elucidate the processes of gliomagenesis with respect to their phenotype, differentiation and tumorigenic capacity during initiation and progression. Research on GSCs is still in its infancy, so no definitive conclusions about their role can yet be drawn. To understand the biology of GSCs fully, it is highly desirable to establish permanent and biologically stable GSC lines. Methods: In the current study, GSCs were isolated from surgical specimens of primary and recurrent glioma in a patient whose malignancy had progressed during the previous six months. The GSCs were cryopreserved and resuscitated periodically during long-term maintenance to establish glioma stem/progenitor cell (GSPC) lines, which were characterized by immunofluorescence, flow cytometry and transmission electronic microscopy. The primary and recurrent GSPC lines were also compared in terms of in vivo tumorigenicity and invasiveness. Molecular genetic differences between the two lines were identified by array-based comparative genomic hybridization and further validated by real-time PCR. Results: Two GSPC lines, SU-1 (primary) and SU-2 (recurrent), were maintained in vitro for more than 44 months and 38 months respectively. Generally, the potentials for proliferation, self-renewal and multi-differentiation remained relatively stable even after a prolonged series of alternating episodes of cryopreservation and resuscitation. Intracranial transplantation of SU-1 cells produced relatively less invasive tumor mass in athymic nude mice, while SU-2 cells led to much more diffuse and aggressive lesions strikingly recapitulated their original tumors. Neither SU-1 nor SU-2 cells reached the terminal differentiation stage under conditions that would induce terminal differentiation in neural stem cells. The differentiation of most of the tumor cells seemed to be blocked at the progenitor cell phase: most of them expressed nestin but only a few co-expressed differentiation markers. Transmission electron microscopy showed that GSCs were at a primitive stage of differentiation with low autophagic activity. Array-based comparative genomic hybridization revealed genetic alterations common to both SU-1 and SU-2, including amplification of the oncogene EGFR and deletion of the tumor suppressor PTEN, while some genetic alterations such as amplification of MTA1 (metastasis associated gene 1) only occurred in SU-2. Conclusion: The GSPC lines SU-1 and SU-2 faithfully retained the characteristics of their original tumors and provide a reliable resource for investigating the mechanisms of formation and recurrence of human gliomas with progressive malignancy. Such investigations may eventually have major impacts on the understanding and treatment of gliomas.