|
|
Genetic
identification of glioblastoma maintaining cells in a xenograft model Christine Glanz, Johan Rebetz, Ylva Stewénius, David Gisselsson, Anna Darabi, Elisabet
Englund, Leif G. Salford, Bengt Widegren, Xiaolong Fan
The Rausing Laboratory, Division of Neurosurgery, Lund University
Hospital, Section of Immunology, Sweden, Department of Clinical
Genetics, Lund, Sweden, The Rausing Laboratory, Division of
Neurosurgery, Lund University Hospital, Lund, Sweden, Department of
Pathology, Lund, Sweden, The Rausing Laboratory, Division of
Neurosurgery, Lund University Hospital, Section of Immunology, Lund
Stem Cell Center, Sweden
Glioblastoma (GBM) can consist
of heterogeneous cell populations.
A small CD133+ stem cell-like
population has been demonstrated to be responsible for GBM
maintenance.
We have developed a serial
transplantable xenograft GBM model to characterize the genetic
background and constitutive signaling pathways in GBM maintaining
cells.
Freshly isolated GBM cells from
a 47-year-old female patient at recurrence were subcutaneously
injected into SCID-Beige mice.
Xenograft GBM formed in initial
SCID-Beige mice were serially transplanted in SCID-Beige or NOD/SCID
mice without cell sorting for up to 4 passages.
Freshly isolated GBM cells and
its xenograft cells at different passages were analyzed for cell
surface marker expression.
Similar patterns of
heterogeneous cell populations were observed both in the GBM and its
xenografts at all passages.
CD44 and epidermal growth
factor receptor were expressed in 90% of the fresh GBM and the
xenograft GBM cells.
The stem cell marker CD133 was
expressed in 70% of the primary GBM and about 30% of the xenograft GBM
cells; the platelet-derived growth factor receptor-a (PDGFRa) in 34%
of the primary GBM and about 20% of the xenograft GBM cells.
However, the immature neural
ganglioside recognized by A2B5 was detected in 63% of the primary GBM
but not in the xenograft GBM cells.
The same pattern of
heterogeneous cell populations was maintained in all passages of
xenograft GBM, suggesting that the GBM cell population hierarchy is to
a great extent maintained in this xenograft model and that a fraction
of GBM cells were capable of maintaining the xenograft GBM by
self-renewal.
G-band karyotyping showed that
the primary GBM cells exhibited great intercellular heterogeneity with
a wide variety of related subclones, all of which showed complex
karyotypes, including several whole-chromosome losses and unbalanced
translocations.
In contrast, late passage
xenograft cells showed almost no intercellular variation and exhibited
a complex karyotype almost identical to one of the original GBM
subclones, identified in 3 out of 25 cells in the primary GBM
cells.
This indicates that the most
subclones in the primary GBM did not contribute to the maintenance of
xenograft GBM.
Interestingly, the PDGFRa
expression in the in-vitro cultured xenograft GBM cells was
specifically down regulated by a 6-day cyclopamine treatment in a
dose-dependent manner.
Concomitantly, a 2,5-fold
reduction of cell proliferation (P < 0,01, n = 3) was observed at
10 μM cyclopamine treatment.
Thus, constitutively active sonic hedgehog signaling critically
contributes to GBM cell proliferation via PDGFRa expression in this
tumor system.
Taken together, our data suggest that even though GBMs consist of
heterogeneous cell populations, only a small fraction of these cells
is responsible for tumor cell regeneration.
(*The first 2 authors contributed equally to this study.)
Copyright © 2005 American
Association for Cancer Research. All rights reserved.
|
