| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Division of Neurological Surgery, Veterans Administration San Diego Health Care System and the University of California-San Diego, San Diego, California 92037
Address all correspondence and requests for reprints to: Hoi Sang U, M.D., Division of Neurosurgery, The Veteran’s Administration Medical Center, 3350 La Jolla Village Drive, La Jolla, California 92037. E-mail: hoisang{at}ucsd.edu.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Initial primary brain cultures were composed of mostly small spindle cells mixed with cells of a fibroblastic and astrocytic morphology. With progressive culture, flat cells declined while the spindle cells predominated. RSCs expressed the nestin message and protein consistent with their CNS progenitor/stem cell identity (Fig. 1
, top and bottom). Expression of the microtubule-associated protein 2 (MAP-2) message was detected at a lower level while the number of MAP-2 immunostaining cells remained rare. The glial fibrillary acidic protein (GFAP) message was not seen and no cell stained for GFAP. Upon removal of basic fibroblast growth factor (bFGF) from the culture medium, the number of nestin+ cells declined while the number of GFAP+ and MAP-2+ cells increased, indicating progressive differentiation into the neurons and glia. This selective differentiation of RSCs along neuronal and glial lineages was maintained throughout culture over a prolonged period (>1 yr). No differentiation of RSCs into other cell types was ever observed.
|
Induction of the Glial Phenotype in RSCs
RSCs are derived from the CNS. We therefore explored whether normal neonatal (P5) and adult astrocytes as well as transformed C6 glial cells could influence their development in cocultures. RSCs were labeled with Bisbenzimide (Sigma, St. Louis, MO) before cocultures. During coculture, GFAP expression was significantly induced in Bisbenzimide+, presumably RSC-derived, cells exposed to developing P5 astrocytes (Fig. 2, A and B), and C6 glioma cells (Fig. 2, C and D). These GFAP+ cells became flattened and extended, processes characteristic of astroglial cells. In these cocultures, expression of nestin was reduced. In contrast, little change in RSCs was noted in cocultures with adult astrocytes (Fig. 2, A and B). Nestin expression was high while GFAP expression was low.
|
|
). However, in view of the very low level of MAP-2 expression in native RSCs, the functional significance of these observations is unclear. In the case of RSCs cocultured with C6 cells, MAP-2 expression was induced (Fig. 2D). Because C6 cells are transformed cells, it is likely that transformation-related properties may be responsible for this induction. Nevertheless, no consistent pattern(s) of MAP-2 expression in these coculture studies was evident.
Therefore, RSCs exposed to developing and neoplastic glial cells were consistently induced to manifest glial properties. In cocultures, differentiating influences may be mediated by cell contact (e.g. connexons) or through the secretion of active substances. To distinguish these two mechanisms, we cultured RSCs for 21 d in media that had been exposed to C6 cells (C6 conditioned medium). In these cultures, RSCs progressively assumed the stellate morphology characteristic of astrocytes with the attendant increase in the number of GFAP+ cells (Fig. 3A), suggesting that factors in the conditioned media induced glial development (Fig. 3B). The expression of GFAP was confirmed by the induction of the GFAP message using RT-PCR.
|
, top). GH3 cells were also immunoreactive with antibodies directed to PRL, human GH (hGH), and Pit-1. Therefore, GH3 cells were remarkably different from RSCs.
|
). The occasional flat cells that showed blue nuclei and stained for nestin are identified as stem cells (Fig. 5A, row 1). There were also, however, some round cells in the cultures that were not positive for Bisbenzimide, suggesting they were GH3 cells.
|
, row 1). None stained for Pit-1, hGH, or PRL. On the other hand, round Bisbenzimide+ cells uniformly stained for Pit-1 (Fig. 5A, row 2), hGH (Fig. 5A, row 3), or PRL (Fig. 5A, row 4), suggesting that they were derived from RSCs that have assumed the morphology, and hGH and Pit-1 expression characteristic of GH3 cells (Fig. 5B). To examine the nature of these transgerm layer induction signals, RSCs were exposed to GH3 conditioned medium.
Upon exposure to GH3 conditioned medium, RSCs did not show any morphological change within the first 2 wk. During this period, expression of the messages for transcription factors, Lhx 3 and Pit-1, which were essential to pituitary development, was evaluated (11, 12, 13, 14, 15, 16, 17, 18) (Fig. 4, bottom). Lhx 3 expression was markedly induced (x20.78 over control) by d 10, whereas Pit-1 expression was barely stimulated (x1.24 over control). Of note, Pit-1 expression was also induced by the serum present in the medium used to maintain GH3 cells, whereas no Pit-1 expression was observed in serum-free medium. By d 15, Lhx 3 expression ceased, being x5.83 below the level seen in cultures exposed to control medium, while Pit-1 expression was initiated (x3.88 control levels). Pit-1 expression was maintained (x2.70 control levels) up to d 20 as RSCs began to assume a more spindle/spherical morphology. The soma of the cells became less spread out on the dish and began to assume a more spindle-like shape. The cellular processes also appeared to contract to form single and shorter protrusions. In longer cultures, the number of spherical cells increased. In these early stages of morphological transformation, spindle-shaped cells began to express selective pituitary hormones characteristic of GH3 cells. These spindle/spherical cells were negative for the nestin protein but expressed Pit-1 (Fig. 5C, row 2), hGH (Fig. 5C, row 3), or PRL (Fig. 5C, row 4, and Fig. 5D). Therefore, RSCs exposed to GH3 conditioned medium, like RSCs in GH3 cocultures, also acquired the morphological and protein expression profiles characteristic of GH3 cells. However, the morphological effects observed in cells exposed to the conditioned media were not as vigorous as those cocultured with GH3 cells and appeared to occur over a longer period of time. In the cultures exposed to GH3 conditioned media, cells that retained their flat morphology remained nestin positive (Fig. 5C, row 1) and did not stain for Pit-1, PRL, or hGH, suggesting nonresponsiveness to the conditioned medium. Thus, one means through which GH3 cells exert their transdifferentiation effects is by the release of soluble factors. This observation was therefore identical to that in RSCs exposed to neonatal and transformed astrocytes. In both situations, RSCs were induced to transdifferentiate in a cell type-specific manner by influences specified by cells derived from two separate germ layers.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
When adult forebrain neural stem cells were injected into the irradiated adult, they generated hematological cells (19). In addition, cells exposed to the primitive environment of the stage 4 chick embryo acquired the phenotypes of many organs generated from the three germ layers (20). In our study, both P5 astrocytes and C6 glioma cells induced CNS stem cells to acquire phenotypes characteristic of astrocytes in cocultures. This finding, together with the failure of adult astrocyte cultures to behave similarly, suggests that these transdifferentiation influences acted through instructive mechanisms instead of permissive mechanisms. In these experimental paradigms, stem cells are subjected to many cell types as well as a host of extracellular signals such as those derived from the extracellular matrix. Both cell-cell contact and the secretion of modulating factors may be responsible for transmitting the transforming signals. In our studies, the induction of astrocytic properties in RSCs by C6 conditioned media demonstrates that the factor(s) responsible for this transdifferentiation may be secreted by the C6 cells and act through paracrine processes. This is supported by similar observations in RSCs exposed to GH3 cells. In RSCs exposed to GH3 conditioned media, the transcription factors, Lhx 3 and Pit-1, essential to pituitary development, were activated in a temporal-specific manner (before the expression of pituitary hormones). This suggests that transdifferentiation was activated through a pituitary-specific transcriptional pathway.
During pituitary development, cell contact as well as secreted factors play important roles in determining cell fate. The anterior pituitary is derived from Rathke’s pouch which is formed as the oral roof ectoderm comes into contact with the overlying neuroepithelium (14, 21). This suggests that initial organ determination is induced by contact-generated signals. This is supported by our coculture experiments. On the other hand, secreted factors such as FGFs and bone morphogen proteins also play an essential role in pituitary cell fate determination (21). Our finding of inductive effects by the GH3 conditioned medium would be consistent with this. However, the morphological effects observed in RSCs exposed to the conditioned media were not as vigorous as those cocultured with GH3 cells and appeared to take a longer period. This suggests that for full transdifferentiation to take place, both contact-mediated events as well as secreted factors were necessary. Nevertheless, our studies suggest that developmental potentials across divides between organs could be extended as long as the specific inductive signals are available. It is likely that once receptive stem cells are exposed to these organ-specific signals, it will activate the appropriate organ-specific transcription factors to ultimately generate the organ-specific phenotype. What is not clear, however, is whether the transdifferentiating stem cells need to revert to a more primitive phenotype(s) before differentiating into the appropriate precursors or that stem cells can directly acquire the new organ-specific phenotypes when appropriately stimulated. In our study, it is demonstrated that GH3 conditioned medium activated in the RSCs pituitary-specific transcription factors such as the Lim homeodomain factor Lhx3 or P-Lim (associated with anterior pituitary development) (16) and the POU-domain gene pit-1 (18) (associated with determination of somatotrophs, lactotrophs, and thyrotrophs), which eventually lead to the generation of specific pituitary phenotypes.
Taken together, these observations indicate that even though the stem cells used in the experiments were all derived from the CNS and thus appeared to be committed to the development of the CNS, they do not seem to be restricted to a defined developmental fate. Therefore, partially committed CNS stem cells may retain pluripotentiality and can be redirected to develop into other cell types not found in the brain, provided the correct set of stimuli is present. In this sense, the developmental potential of stem cells is more universal than previously thought.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Cell Culture
Coculture.
RSCs were cocultured with primary rat neonatal (P5) and adult astrocytes as well as rat C6 glioma and GH3 pituitary adenoma cells. For coculture with primary rat neonatal (P5) and adult astrocytes, these cells were first isolated from the respective animals. Brains from P5 pups and adult Sprague Dawley rats were harvested, filtered to generate a cell suspension, and plated in culture (22). After 7 d, the dishes were agitated and rinsed to remove the less adherent neurons and oligodendrocytes. The resultant culture was composed of more than 98% GFAP+ astrocytes.
For coculture, RSCs were first labeled for 3 d with 20 µM Bisbenzimide (Sigma), which binds to DNA and fluoresces under an Ultraviolet filter. This allowed us to identify cells of RSC origin as Bisbenzimide+.
On the day of coculture, 105 Bisbenzimide-labeled RSCs were plated onto the astrocyte, C6 or GH3 (ATCC, Manassas, VA) cultures on poly-L-ornithine-coated coverslips in DMEM + 10% FCS. RSCs were first analyzed immediately after initial plating (d 0) to provide a baseline characterization of cell marker expression. After 2 wk, the samples were processed for immunocytochemistry analysis of the expression of nestin and GFAP, or nestin and pituitary-related factors. As a negative control to evaluate GFAP or pituitary-specific hormone expression, RSCs not exposed to C6 cells or GH3 cells were used. For positive control, C6 cells or GH3 cells not cocultured with RSCs were used. In these sets of cocultures, three kinds of media were also used to control for the effects of the respective media: 1) DMEM/F12 supplemented with N2 but not with any growth factor, 2) the Ham’s/F12 + 15% horse serum (HS) + 2.5% FCS medium used to maintain GH3 cells, and 3) DMEM + 10% FCS medium used to maintain C6 cells.
RSC Cultures Exposed to Conditioned Media
Media exposed to C6 glioma (DMEM + 10% FCS) or GH3 (Ham’s/F12 + 15% HS + 2.5% FCS) cells were collected every 6 d and immediately filtered (0.2 µm filter) without any further processing. Before use on the RSC cultures, the media was diluted 1:1 with DMEM/F12 medium supplemented with N2. To induce the RSCs, conditioned medium was added to each RSC culture maintained on poly-L-ornithine-coated coverslips and dishes and changed every 3 d. After 20 d of conditioning, RSC cultures were fixed for immunocytochemical analysis.
Characterization of Nestin, MAP-2, GFAP, Lhx-3, Pit-1, and PRL Expression Using Quantitative RT-PCR in RSC Cells
RSCs were seeded in duplicate at approximately 1 x 105 cells per 60-mm tissue culture plate and evaluated for the expression of nestin, MAP-2, and GFAP at the message level.
RNA was extracted from cells using Trizol (Life Technologies, Inc.). Three micrograms of RNA were reverse transcribed into cDNA using the Superscript II Preamplification System (Life Technologies, Inc.). Quantitative PCR was conducted to assay the level of GFAP message using glyceraldehyde-3-phosphate dehydrogenase (GAPDH) as an internal control. The PCR (volume 25 µl) included: 1x PCR buffer (Life Technologies, Inc.), 2 mM MgCl2 (Life Technologies, Inc.), 0.4 mM deoxynucleotide triphosphates (Life Technologies, Inc.), 0.2 µM oligo primers, 0.5 µl of the reverse transcription product, and 1.5 U Amplitaq (Perkin-Elmer Corp.). The PCR sequence was: 95 C for 3 min, 35 cycles of reaction at 94 C for 1 min, 54 C for 1 min, 72 C for 2 min, and then 72 C for 10 min. The primers used were selected for rat nestin, MAP-2, GFAP, GAPDH (Life Technologies, Inc.), Lhx3, Pit-1, and PRL:
The RT-PCR products were analyzed in a 2% agarose gel after staining with ethidium bromide. The size of each band for each specific PCR product was then determined using a 50-bp ladder to confirm that they represented the correct product as designed from the specific primers. As control, rat pancreatic tissues were used. The band size for each PCR product was: nestin, 249 bp; MAP-2, 213 bp; GFAP, 289 bp; Pit-1, 219 bp; PRL, 182 bp, Lhx3, 239 bp. In this analysis, the presence or absence of a specific band would indicate the expression of each specific message. Where appropriate, the intensity of each band in control and experimental samples was determined and compared as an estimation of the degree of gene induction. To this end, the NIH image software was used to analyze band intensity in photographs of the electrophoretic gels comparing the RT-PCR products at each time point. Each band in the gel photograph was scanned and plotted according to intensity. The area under each curve representing a band was measured and used for comparison analysis.
Immunocytochemical Characterization of the Expression of Cell Type-Specific Markers
The RSC/Astrocyte, RSC/C6 Glioma, and RSC/GH3 Cocultures.
Cells on glass coverslips were fixed with 4% paraformaldehyde (in PBS) for 1 h at 22 C, exposed to Triton X-100 (0.5% in PBS) for 10 min, and treated with blocking buffer (5% normal goat serum in PBS) for 30 min at 22 C. For characterization of the natural development of isolated RSCs as well as their development upon exposure to glial cells, cultures were reacted with one of the following primary antibodies: 1) a mouse monoclonal antibody against nestin at 1:500 dilution (PharMingen, San Diego, CA); 2) a rabbit polyclonal antibody specific for bovine GFAP at 1:200 dilution (DAKO Corp., Carpinteria, CA); or 3) a mouse monoclonal antibody specific for MAP-2 at 1:200 dilution (PharMingen). Controls consisted of staining with PBS/5% normal goat serum from which the primary antibodies were omitted as well as preimmune serum. For characterization of the expression of pituitary factors and hormones, cultures were exposed to one of the following primary antibodies: 1) a mouse monoclonal antibody against nestin at 1:500 dilution (PharMingen); 2) a goat anti-PRL antibody at 1:200 dilution (Santa Cruz Biotechnology, Inc., Santa Cruz, CA); 3) a rabbit anti-hGH antibody (DAKO Corp.) at 1:400 dilution; or 4) a rabbit anti-Pit 1 antibody at 1:200 dilution (Santa Cruz Biotechnology, Inc.).
After 1 h at 37 C, the cells were washed extensively with PBS. Cells were then reacted for 30 min at 37 C with a second antibody which is either 1) a goat antirabbit IgG conjugated to fluorescein (1:100 dilution in PBS/5% normal goat serum) (Sigma); or 2) a goat antimouse IgG conjugated to rhodamine (1:25 dilution; Sigma); or 3) rhodamine-conjugated goat antirabbit IgG (1:80 dilution); or 4) rhodamine-conjugated rabbit antigoat IgG (1:80).
In this analysis, RSCs were first identified by viewing the samples using a UV filter which revealed the Bisbenzimide-labeled RSC nuclei as an intense light blue-stained structure. With the same view in place, the morphology and the expression of each specific factor were recorded for RSC-derived cells. At least 5–10 random high power fields consisting of greater than 50 cells are examined under each condition for each cell marker. The percentage of cells in each field expressing a specific marker was then determined. Comparisons between control and experimental groups were made using Student’s t test. Significant differences (P < 0.05) were indicated with an asterisk.
In these coculture studies, the Bisbenzimide label could occasionally be released from labeled cells (e.g. when the labeled cells die). This could potentially be taken up by the surrounding non-RSC-derived cells and present as artefactual labeling. We therefore have conducted the control experiment where media exposed to labeled RSCs were collected using the schedule detailed in the conditioned media experiments. Media exposed to Bisbenzimide-labeled cells were then transferred to cultures not previously exposed to Bisbenzimide. After 2 wk, the treated cultures were examined for the uptake of Bisbenzimide. Even though some uptake of Bisbenzimide was observed, this occurred only in random patches of cells (at most <30% of the total number of cells in the dish). However, the degree of uptake was generally low and was at a level slightly above background. For this reason, the scoring only of those cells with nuclear Bisbenzimide labeling at least 5- to 10-fold above background levels was essential to render the analysis meaningful.
RSC Cultures Exposed to Glial and GH3 Conditioned Media.
These samples were similarly examined as above with the exception that identification of RSCs was not necessary.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Abbreviations: bFGF, Basic fibroblast growth factor; CNS, central nervous system; E12, embryonic d 12; FCS, fetal calf serum; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; GFAP, glial fibrillary acidic protein; hGH, human GH; HS, horse serum; MAP-2, microtubule-associated protein 2; P5, postnatal d 5; PRL, prolactin; RSC, rat fetal CNS stem cells.
Received for publication March 7, 2002. Accepted for publication July 26, 2002.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  O. Krylyshkina, J. Chen, L. Mebis, C. Denef, and H. Vankelecom Nestin-Immunoreactive Cells in Rat Pituitary Are neither Hormonal nor Typical Folliculo-Stellate Cells Endocrinology, May 1, 2005; 146(5): 2376 - 2387. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. J Burns, S. J Persaud, and P. M Jones Stem cell therapy for diabetes: do we need to make beta cells? J. Endocrinol., December 1, 2004; 183(3): 437 - 443. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
