| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
|
FJ
EXPRESS SUMMARY ARTICLE The Full-length version of this article is also available, published online October 14, 2004 as doi:10.1096/fj.04-1931fje. |
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Institute of Reconstructive Neurobiology, University of Bonn and Hertie Foundation, Bonn, Germany
1Correspondence: Institute of Reconstructive Neurobiology, University of Bonn Medical Center, Sigmund Freud Strasse 25, Bonn 53105, Germany. E-mail: brustle{at}uni-bonn.de
SPECIFIC AIMS
In many central nervous system (CNS) disorders, including multiple sclerosis, spinal cord injury, and stroke, demyelination of intact axons is an important factor contributing to loss of function. Transplantation of myelin-forming cells into the CNS has become the most promising experimental strategy for restoring lost myelin. Availability of embryonic stem (ES) cells has provided new prospects for generating tissue-specific stem and progenitor cells in virtually unlimited numbers in vitro. Due to their pluripotency, their potential for limitless self-renewal, and their amenability to genetic modification and gene targeting, ES cells are widely considered as the most versatile donor source for cell and tissue repair. The therapeutic application of ES cells for myelin repair will depend critically on the ability to enrich a pure population of mitotically active oligodendrocyte progenitors (OPs) able to differentiate into myelinating oligodendrocytes.
In this study, we explore a new strategy for the enrichment and isolation of ES cell-derived OPs. Growth factor-controlled differentiation of ES cells into bipotential glial precursors was combined with a subsequent oligodendrocyte-specific lineage selection step. Using a selection construct carrying a neomycine resistance gene under control of the murine oligodendrocyte-specific CNP promoters I and II (CNP-ßgeo), we demonstrate that this strategy can be used to generate highly purified cultures of proliferating OPs capable of myelin formation in vivo.
PRINCIPAL FINDINGS
1. Efficient oligodendroglial lineage selection using the CNP-ßgeo transgene
To provide a basis for lineage selection, CNP-ßgeo-engineered mouse ES cell clones were subjected to controlled in vitro differentiation into bipotential glial precursors (Fig. 1
). After aggregation to embryoid bodies, spontaneously differentiating neuroepithelial cells were selected in a defined medium containing insulin, transferrin, selenium, and fibronectin (ITSFn). Multipotent neural precursors were proliferated in the presence of FGF-2, and the cells were subsequently shifted into a glial precursor fate by sequential propagation in FGF-2/EGF and FGF-2/PDGF. After growth factor-induced differentiation, up to 40% of the glial precursors derived from the transfected ES cell clones showed ß-gal staining. Glial precursors derived from 9 clones with the highest ßgal expression levels were further subjected to neomycine selection for CNP-ßgeo-expressing cells. To enable efficient lineage selection, we developed a specific sequence of culture conditions which induce and stabilize a CNP-positive fate while avoiding postmitotic terminal differentiation (Fig. 1)
. To that end, glial precursors proliferating in the presence of FGF-2 and PDGF were subjected to a temporary, 4 day growth factor withdrawal in the presence of T3, a hormone known to facilitate the differentiation of oligodendrocytes. This predifferentiation step resulted in a cell population encompassing small, process-bearing cells with condensed cell bodies and flat cells with an astrocyte-like morphology. These cells were further propagated for 2 days in serum-containing medium in the presence of FGF-2, PDGF, and T3. The cultures were subjected to neomycine selection in the presence of FGF-2 and PDGF for an additional 2–3 days. Most of the cells with flat, astrocytic morphology died, leaving a homogenous population of bipolar cells with poorly branched processes characteristic of OPs. Glial precursors derived from wild-type ES cells did not survive this purification process, confirming the specificity of the CNP-ßgeo-based selection. Immunofluorescence analysis showed that 96 ± 7% of the selected cells coexpress CNP and ß-gal (Fig. 2A, B
). These data demonstrate that the CNP-ßgeo transgene permits efficient selection of CNP-expressing cells from differentiating ES cells.
|
|
2. CNP promoter-based lineage selection yields a mitotically active population with a pre-oligodendrocyte phenotype
We next examined the developmental stage and the differentiation potential of the selected cell population. Quantitative results of the antigenic marker expression are summarized in Fig. 2A
. Of the cells obtained by CNP-based lineage selection, 80 ± 7% still displayed nestin immunoreactivity (Fig. 2A, E
). Nestin has been shown to be expressed in proliferating OPs but is down-regulated in differentiated oligodendrocytes. A large fraction (88±9%) of the cells expressed the polysialylated or embryonic form of the neural cell adhesion molecule NCAM (PSA-NCAM) (Fig. 2A, C
). PSA-NCAM is considered to define the precursor stage from which oligodendrocytes arise, although it is also expressed in other neural and non-neural cell types. In addition to the expression of nestin and PSA-NCAM, 95 ± 6% of the cells displayed immunoreactivity to the A2B5 antibody, which recognizes gangliosides expressed by dividing OPs in vitro (Fig. 2A, D
). A fraction of 78 ± 10% was found to be positive for the O4 antigen (Fig. 2A, G
), indicating incipient oligodendroglial differentiation. Only a small subset (10±6%) of the selected cells expressed NG-2 (Fig. 2A, F
), an antigen typically detected in OPs. A low rate of spontaneous differentiation into more mature oligodendrocytes was suggested by the presence of GalC-positive cells (Fig. 2A
; 9±4%). A small fraction of the selected cells were found to express GFAP (Fig. 2A
; 3±3%), a marker not only expressed in astrocytes but also in immature and nonmyelinating Schwann cells (SCs). Since CNP is expressed in SCs, we studied the expression of the transcription factor Krox20 and the myelin protein P0 (i.e., markers characteristic of a SC fate). Immunocytochemical analysis using Krox20- and P0-specific antibodies failed to detect any immunoreactive cells among the selected population. Western blot analyses confirmed the absence of Krox20 whereas CNP and the ßgeo fusion protein were strongly expressed in the selected cells (Fig. 2H
). These observations indicate that the in vitro differentiation protocol employed does not provide sufficient precursors for a CNP-ßgeo-based selection of ES cell-derived SCs. Thus, the GFAP-positive population most likely represents occasional astrocytes which had escaped neomycine selection. Immunocytochemical studies with an antibody to ß-III tubulin confirmed the absence of neurons in selected cell populations. Taken together, the coexpression of CNP with nestin, PSA-NCAM and A2B5 (i.e., markers typically associated with a neural precursor stage) indicates that the majority of the selected cells represents an early pre-oligodendrocyte phenotype.
Analysis of BrdU incorporation in the presence of the mitogens FGF-2 and PDGF demonstrated that the selected cells continued to proliferate with a doubling time of 
72 h (Fig. 1
, left panel), confirming that they still represent precursors rather than postmitotic oligodendrocytes. Growth factor withdrawal for three days induced differentiation into a more mature, GalC-positive phenotype in 80 ± 7% of the population (Fig. 1
, right panel). These data demonstrate that CNP promoter-based lineage selection of in vitro predifferentiated ES cells yields a population of mitotic OPs amenable to FGF-2/PDGF-mediated proliferation and growth factor withdrawal-induced differentiation.
3. The selected cell population exhibits a myelinating phenotype in vivo
To study whether the selected ES cell-derived OPs can, in principle, generate myelinating oligodendrocytes in vivo, we performed a first set of transplant experiments in the spinal cord of 4–6-day-old myelin-deficient(md) rats. This mutant develops severe CNS hypomyelination due to a point mutation in the X-linked PLP gene. The phenotype of these rats resembles the human myelin disorder Pelizaeus Merzbacher disease, which is also associated with aberrations in the PLP gene. Due to the lack of endogenous myelin formation and the absence of PLP expression, donor-derived internodes can be easily detected by PLP immunolabeling. Two weeks after transplantation, the engrafted cells had formed PLP-positive myelin internodes (Fig. 1
, lower panel). Double labeling of PLP-immunolabeled cells by fluorescence DNA in situ hybridization with a mouse-specific probe confirmed the donor origin of the engrafted cells and thus demonstrated the myelinogenic potential of the purified cell population.
CONCLUSIONS AND SIGNIFICANCE
The results of this study demonstrate that the combination of growth factor-controlled ES cell differentiation and CNP promoter-based lineage selection permits the generation of highly purified OPs. The purified cell populations remain proliferative in vitro and, upon transplantation, generate myelinating oligodendrocytes in vivo. ES cell-derived OPs generated with this method may serve as an attractive donor source for pharmacological and toxicological studies, compound development, and myelin repair.
FOOTNOTES
To read the full text of this article, go to http://www.fasebj.org/cgi/doi/10.1096/fj.04-1931fje;
This article has been cited by other articles:
|
|
 |
|
  A. K. Shetty and B. Hattiangady Concise Review: Prospects of Stem Cell Therapy for Temporal Lobe Epilepsy Stem Cells, October 1, 2007; 25(10): 2396 - 2407. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Haupt, F. Edenhofer, M. Peitz, A. Leinhaas, and O. Brustle Stage-Specific Conditional Mutagenesis in Mouse Embryonic Stem Cell-Derived Neural Cells and Postmitotic Neurons by Direct Delivery of Biologically Active Cre Recombinase Stem Cells, January 1, 2007; 25(1): 181 - 188. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Piepoli, C. Jakupoglu, W. Gu, E. Lualdi, B. Suarez-Merino, P. L. Poliani, M. G. Cattaneo, B. Ortino, D. Goplen, J. Wang, et al. Expression studies in gliomas and glial cells do not support a tumor suppressor role for LGI1 Neuro-oncol, April 1, 2006; 8(2): 96 - 108. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
150 kDa) in lineage selected cultures derived from two independent CNP-ßgeo ES cell clones (#9, #21). Note the increase in CNP expression as compared with an unselected wild-type ES cell-derived astrocyte/oligodendrocyte culture (A/O). The Schwann cell marker Krox20 (48 kDa) is detected in extracts from sciatic (SN) and brachial (BN) nerve but absent in CNP-ßgeo-selected cells.