Stem Cells
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

First published online December 20, 2007
Stem Cells Vol. 26 No. 3 March 2008, pp. 656 -665
doi:10.1634/stemcells.2007-0810; www.StemCells.com
© 2008 AlphaMed Press

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Data
Right arrow All Versions of this Article:
2007-0810v1
26/3/656    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Reprints/Permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Flanagan, L. A.
Right arrow Articles by Monuki, E. S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Flanagan, L. A.
Right arrow Articles by Monuki, E. S.

TISSUE-SPECIFIC STEM CELLS

Unique Dielectric Properties Distinguish Stem Cells and Their Differentiated Progeny

Lisa A. Flanagana, Jente Lua, Lisen Wangb, Steve A. Marchenkoa, Noo Li Jeonb, Abraham P. Leeb, Edwin S. Monukia,c

Departments of aPathology and Laboratory Medicine, School of Medicine,
bBiomedical Engineering, Henry Samueli School of Engineering, and
cDevelopmental and Cell Biology, School of Biological Sciences, University of California Irvine, Irvine, California, USA

Key Words. Neural stem cell • Cerebral cortex • Cortical • Sorting • Dielectrophoresis • Progenitor

Correspondence: Correspondence: Lisa A. Flanagan, Ph.D., Department of Pathology and Laboratory Medicine, Medical Sciences I, D-440, University of California Irvine School of Medicine, Irvine, California 92697-4800, USA. Telephone: 949-824-5786; Fax: 949-824-2160; e-mail: lflanaga{at}uci.edu; or Edwin S. Monuki, M.D., Ph.D., Department of Pathology, Department of Pathology, Medical Sciences I, D-440, University of California Irvine School of Medicine, Irvine, California 92697-4800, USA. Telephone: 949-824-9604; Fax: 949-824-2160; e-mail: emonuki{at}uci.edu

Received on October 8, 2007; accepted for publication on December 4, 2007.

First published online in STEM CELLS EXPRESS  December 20, 2007.

   ABSTRACT
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 Disclosure of Potential...
 Acknowledgments
 References
 
The relatively new field of stem cell biology is hampered by a lack of sufficient means to accurately determine the phenotype of cells. Cell-type-specific markers, such as cell surface proteins used for flow cytometry or fluorescence-activated cell sorting, are limited and often recognize multiple members of a stem cell lineage. We sought to develop a complementary approach that would be less dependent on the identification of particular markers for the subpopulations of cells and would instead measure their overall character. We tested whether a microfluidic system using dielectrophoresis (DEP), which induces a frequency-dependent dipole in cells, would be useful for characterizing stem cells and their differentiated progeny. We found that populations of mouse neural stem/precursor cells (NSPCs), differentiated neurons, and differentiated astrocytes had different dielectric properties revealed by DEP. By isolating NSPCs from developmental ages at which they are more likely to generate neurons, or astrocytes, we were able to show that a shift in dielectric property reflecting their fate bias precedes detectable marker expression in these cells and identifies specific progenitor populations. In addition, experimental data and mathematical modeling suggest that DEP curve parameters can indicate cell heterogeneity in mixed cultures. These findings provide evidence for a whole cell property that reflects stem cell fate bias and establish DEP as a tool with unique capabilities for interrogating, characterizing, and sorting stem cells.

Disclosure of potential conflicts of interest is found at the end of this article.


   INTRODUCTION
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 Disclosure of Potential...
 Acknowledgments
 References
 
A current limitation in the stem cell field is the inability to accurately identify or separate stem cells from more differentiated progeny. During the differentiation process, stem cells form sequentially more specified progenitor cells that eventually generate differentiated daughter cells. "Stem" cells are able to self-renew over a long period and generate all the cell types found in a tissue, whereas "progenitor" cells are more limited in both proliferative potential and the array of differentiated cell types they can form. In many cases, clear identification of particular progenitor cell populations would enable studies of stem cell biology or provide a means to generate specific cell types for therapy. For example, neural stem cells of the cerebral cortex predominantly give rise to neurons early in development but form glia at later stages, suggesting changes in progenitor cell fate potential over time [1–6]. Further characterization of neuron-restricted progenitors (or their use in transplantation experiments) has been hindered by the lack of specific markers that will distinguish these progenitors from both multipotent stem cells and more differentiated cells. With no means to specifically distinguish one set of cells from the other, purifying or recognizing a particular subset of cells for study or therapeutic uses is impossible. Furthermore, expansion of stem cell populations in vitro, as would be necessary to generate sufficient numbers of cells for transplantation, is often accompanied by differentiation of subsets of cells in the culture [7], leading to heterogeneity in the population. Without ways to discriminate and isolate subpopulations of cells in the culture, controlling the composition of cell transplants is problematic.

Identification of subtypes of cells in a population has traditionally been accomplished by flow cytometry or fluorescence-activated cell sorting (FACS). These techniques rely upon the availability of cell-surface antigens that are unique to the cell type of interest and specific antibodies that can detect these antigens. This approach has been used successfully in the hematopoietic stem cell field and often uses a combination of the presence of particular markers and the absence of others. However, for many stem cell populations, cell surface markers for this type of analysis are severely limited or lacking altogether.

We sought to determine whether stem cells and their more differentiated progeny could be identified by other means. In particular, we hoped to identify a nonbiased approach that would probe characteristics of the entire cell and not require the expression of a certain set of markers on the cell surface. We chose to test whether the overall "electrical signatures" of the cells could be used to identify stem and more differentiated cells. Cells have previously been distinguished from one another using this signature in a method termed dielectrophoresis (DEP), which uses a nontoxic electrical stimulation to induce a frequency-dependent dipole in cells [8]. DEP detects inherent cell traits such as surface charge, membrane conductivity and structure, nucleic acid content, cell size, and presence and conductivity of internal membrane-bound vesicles and charged cytoplasmic molecules [9]. DEP has been used to distinguish similar populations of cells, including stimulated versus nonstimulated Jurkat cells, subpopulations of human leukocytes, breast cancer cells transfected with the neu oncogene, and neurons from astrocytes [10–16]. With the exception of the enrichment of CD34+ hematopoietic stem cells from bone marrow or peripheral blood [17, 18], DEP has not been used to characterize stem cells and their differentiated progeny.

We measured the DEP responses of mouse neural stem/precursor cells (NSPCs) and their differentiated derivatives (neurons and glia) and found that these cells have distinct dielectric properties. Moreover, we determined that DEP signatures distinguish NSPCs from different developmental ages in a fashion that predicts their respective fate biases. These data suggest that the developmental progression of progenitor cell populations can be revealed by the cells' dielectric properties and also show the sensitivity of DEP to minor changes in cell phenotype. We further found that DEP can be used to indicate the heterogeneity of a population of cells, providing another measure for characterizing stem cell cultures.


   MATERIALS AND METHODS
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 Disclosure of Potential...
 Acknowledgments
 References
 
Cells
Mouse fetal-derived NSPCs were cultured from cerebral cortical regions of wild-type CD1 mice at embryonic days 12.5 and 16.5 (E12.5 and E16.5) [19]. Cultures of NSPCs were grown as neurospheres in Dulbecco's modified Eagle's medium, B27, N2, 1 mM sodium pyruvate, 2 mM glutamine, 1 mM N-acetylcysteine (Sigma-Aldrich, St. Louis, http://www.sigmaaldrich.com), 20 ng/ml epidermal growth factor (BD Biosciences, Bedford, MA, http://www.bdbiosciences.com), 10 ng/ml fibroblast growth factor (BD Biosciences), and 2 µg/ml heparin (Sigma-Aldrich) (all culture reagents from Gibco [Grand Island, NY, http://www.invitrogen.com] unless otherwise specified). Neurospheres are a heterogeneous collection of cells that includes a small number of stem cells, a greater number of more specified progenitor cells, and a few differentiated cells [7]. For differentiation, neurospheres were dissociated, and cells were plated on laminin-coated coverslips in medium lacking growth factors and heparin. Neurons and astrocytes from E12.5 mouse cortices were cultured on coverslips coated with Matrigel (BD Biosciences) using medium described previously [20]. Neurons were cultured for 2 days prior to analysis. Astrocytes were generated by passaging the cells with trypsin-EDTA several times over the course of ~14 days to enrich for astrocytes and remove neurons. In most cases, cultures of NSPCs, neurons, and astrocytes were derived from the same set of embryos.

NSPC neurospheres were dissociated for DEP assays using NeuroCult dissociation buffer (StemCell Technologies, Vancouver, BC, Canada, http://www.stemcell.com), and cells were resuspended in DEP buffer (8.5% sucrose [wt/vol], 0.3% glucose [wt/vol], and 0.725% [vol/vol] RPMI) of conductivity 150 µS/cm as measured by a conductivity meter (Thermo Orion, Beverly, MA, http://www.thermo.com). To control for different culture conditions and means of dissociation, NSPCs were also plated as adherent cells on Matrigel-coated coverslips, cultured for 2 days, and harvested by brief (≤5 minutes) trypsinization prior to DEP testing. NSPC DEP profiles were similar for cells from both conditions (data not shown). Differentiated cells were harvested for DEP assays by brief trypsinization after 2 days (neurons) or 2 weeks (astrocytes) in vitro and resuspended in DEP buffer. Mouse neuroblastoma cells (N115) were cultured in Eagle's minimum essential medium (EMEM) (Mediatech, Manassas, VA, http://www.cellgro.com) containing 0.6% glucose (wt/vol) and 10% fetal bovine serum (FBS). Human embryonic kidney cells (293 cells) were grown in EMEM containing 10% FBS. Cell cultures were maintained in a humidified tissue culture incubator at 37°C and 5% CO2. Cell counts and viability were determined for all cultures using trypan blue exclusion. Viability of cells in the DEP buffer was assessed after incubating the cells in suspension at room temperature for 1–6 hours. Immunostaining was as previously described [19] and used the following antibodies: anti-glial fibrillary acidic protein (anti-GFAP) polyclonal, 1:1,000 (Chemicon, Temecula, CA, http://www.chemicon.com), or monoclonal, 1:400 (Sigma-Aldrich); anti-MAP2 (microtubule-associated protein 2) (HM2) monoclonal, 1:100 (Sigma-Aldrich); anti-Nestin (rat 401) monoclonal, 1:1,000 (Developmental Studies Hybridoma Bank [Iowa City, IA, http://www.uiowa.edu/~dshbwww], developed under the auspices of the National Institute of Child Health and Human Development); and anti-Sox2 (Y17) polyclonal, 1:200 (Santa Cruz Biotechnology Inc., Santa Cruz, CA, http://www.scbt.com).

DEP Device Fabrication
DEP devices were fabricated on either silicon wafers or glass slides using modifications of techniques previously described [21] (additional details given in supplemental online data). The dimensions of the DEP microchannels were 1.8–2.0 cm in length, 500 µm in width, and 50 µm in height. There were three groups of electrodes in each channel, and each group contained 16 electrodes with widths of 50 µm and gaps of 100 µm between electrodes.

Cell Trapping in the DEP Device
For cell trapping experiments, the DEP device was placed on the stage of an upright microscope (Nikon Eclipse L150; Tokyo, http://www.nikon.com), and the channel region of the device was visualized with a 10x objective. The channel was cleaned by trypsin-EDTA followed by extensive washing with DEP buffer prior to loading cells. The microchannel outlet of the DEP device was connected to a syringe pump (Pico Plus; Harvard Apparatus, Holliston, MA, http://www.harvardapparatus.com) by Teflon tubing. Fluid flow was initiated in the channel by filling the microchannel inlet with DEP buffer and pumping at a high flow rate (40 µl/minute) from the outlet until fluid completely filled the channel and attached tubing. To stabilize fluid flow, DEP buffer was equilibrated through the channel for 10 minutes at the flow rate to be used in the experiment before loading cells into the channel. Cells were loaded into the channel at a density of 8 x 105 cells per milliliter in DEP buffer, with flow rates ranging from 0.1 to 10 µl/minute. A function generator (Tektronix AFG320; Tucker, Garland, TX, http://www.tucker.com) was used to produce an AC signal (8 V; frequency varied from 25 kHz to 10 MHz) to provide the DEP force in the channel. Cells flowed through the channel for ~5 seconds before the DEP force was applied. A color digital camera (Spot Insight QE; Diagnostic Instruments, Sterling Heights, MI, http://www.diaginc.com) attached to the microscope and CamStudio screen recorder were used to capture ~10–15-second movies (100 frames per second) of the cells trapped at each frequency. Cell trapping data were collected in three to five separate experiments, and two or three video segments were captured for each frequency in each experiment.

Data Analysis
The percentage of trapped cells at each frequency was determined by viewing videos (Photron Fastcam Viewer; Photron, San Diego, http://www.photron.com) and manually counting the cells flowing through the channel in a 5-second time frame (~50–100 cells per 5-second video). The cells trapped along the electrode arrays were expressed as a percentage of the total number of visible cells. Statistical analyses used Student's t test, and p < .05 was considered significant. All data are expressed as means ± SEM of at least three independent experiments, and more than 450 cells were counted at each frequency.

Hydrodynamic Force and DEP Simulations
For the hydrodynamic force simulation, a finite element computer program (FEMLAB 3.1; COMSOL, Los Angeles, http://www.comsol.com) was used to simulate the three-dimensional velocity profile in the channel and to calculate the shear stress at the bottom of the channel (additional details given in supplemental online data). The DEP simulation used a MATLAB program (version 7; MathWorks, Natick, MA, http://www.mathworks.com) and DEP force equations [8, 22] to generate trapping efficiency curves for hypothetical homogeneous cell populations (additional details given in supplemental online data). Trapping efficiency curves were then generated for a hypothetical heterogeneous population of cells composed of various percentages of the homogeneous cells.


   RESULTS
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 Disclosure of Potential...
 Acknowledgments
 References
 
Design and Testing of the DEP Device
The experimental setup used to characterize cells by DEP is illustrated in Figure 1A. The DEP device is a fabricated microfluidic platform (details given in Materials and Methods) that is compatible with microscopy. The DEP device, upright microscope for visualization of cells, syringe pump to control fluid flow, function generator to provide the DEP force, and camera connected to a computer to record videos to document cell behavior constitute the complete experimental setup. The DEP device itself is composed of a microfluidic channel, inlet and outlet ports for the channel, and electrodes that lie on the floor of the channel and provide the DEP force. When an appropriate frequency is applied to the electrodes, cells in the channel are trapped along the electrode sides (Fig. 1B) because of the distribution of the DEP force along the electrode [21]. We previously used simulation programs to inform the electrode design and determined that 50-µm electrodes spaced 100 µm apart would provide a strong and well-spaced DEP electric field [21]. The response of cells to DEP can be characterized by exposing the cells to an electrical signal of a particular voltage and a range of frequencies. At lower frequencies, the cells will be repelled from the electrodes (negative DEP), and at higher frequencies, the cells will be attracted to the electrodes (positive DEP). Different cells experience positive DEP at different frequencies depending on their composite electrical properties [9]. Therefore, at a particular frequency, one set of cells may be attracted to the electrodes by positive DEP, whereas a different set will be repelled by negative DEP (Fig. 1C).


Figure 1
View larger version (77K):
[in this window]
[in a new window]

  		Figure 1. DEP setup, device, and verification. (A): The experimental setup consisted of a DEP device (1) that was placed on the stage of an upright microscope and connected to a syringe pump (2) via the device outlet to control fluid flow. The device electrodes were connected to a dual channel function generator (3) to produce the electrical force. Cells were loaded in the device and then monitored using the upright microscope (4) and color camera (5) connected to a computer (electrodes in the device channel are shown on computer screen). (B): Image of the DEP device. The channel inlet and outlet are marked by black circles, and electrode connectors are visible at the edges of the slide. The area outlined by the blue rectangle is schematized below to show the channel inlet (the dark circle is a hole in the PDMS that allows access to the inlet), channel, and interdigitated electrodes along the channel bottom. The red-boxed region is enlarged in the top right panel to show a higher-magnification view of the electrodes in the channel (the electrodes appear white in the higher magnification image). The bottom right panel shows a higher-magnification view of the electrodes when the DEP force is applied and cells are trapped. (C): The top panel schematic depicts the differential responses of two unique hypothetical cell types to DEP forces (schematic modified from [13]). Traces on the graph show responses of the cells to different electrical frequencies. At a particular frequency (denoted by the dashed green line), cells move either toward the electrode (positive DEP) or away from the electrode (negative DEP). The lower panel shows the differential DEP responses of two cell lines. Human embryonic kidney (293 cells) exhibited positive DEP at ~350 kHz, whereas mouse neuroblastoma cells (N115 cells) experienced positive DEP at ~900 kHz. Abbreviation: DEP, dielectrophoresis; PDMS, poly-dimethylsiloxane.

 
We initially tested the fidelity of our DEP system by assessing the dielectric properties of two distinct cell lines: human embryonic kidney cells (293 cells) and mouse neuroblastoma cells (N115 cells). In order for cells to experience the DEP force, the conductivity of the surrounding buffer must be low so that an electric dipole can be induced in the cell and not screened by charged particles in the buffer. We determined that a sucrose-based buffer with a conductivity of 150 µS/cm was compatible with these cells and allowed sufficient transmittance of the DEP force in our devices. We screened cell responses to electrical frequencies ranging from 25 kHz to 10 MHz and determined that frequencies that elicited positive DEP from 293 cells were much lower than the frequencies necessary to induce positive DEP of N115 cells (Fig. 1C). These results show that our DEP system can easily detect the unique dielectric properties of distinct populations of cells.

Compatibility of Neural Cells with the DEP Device
We used established protocols to culture undifferentiated fetal-derived NSPCs, differentiated neurons, or differentiated astrocytes from the embryonic mouse cortex and confirmed their identity using standard morphologic and immunophenotypic criteria (Fig. 2A; [19, 20, 23, 24]). Some NSPCs in late embryonic and postnatal cortical regions express GFAP [25–29], resulting in confusion about the use of GFAP as a marker for astrocytes. The GFAP-positive cells in our studies were astrocytes and not NSPCs based on the following criteria: (a) they were isolated from an embryonic stage (E12.5) at which NSPCs do not express GFAP [26], (b) they did not coexpress NSPC markers such as Nestin [27, 30], and (c) they were flat cells with characteristic stellate or spread astrocyte morphology [27, 30]. We tested whether the sucrose-based DEP buffer would support the viability of all three cell types over the time frame necessary for DEP experiments (less than 3 hours) and found no loss of viability of any of the cells after 6 hours (the latest time point tested) (Fig. 2B). These results show that cell death due to prolonged incubation in the DEP buffer should not be a complicating factor in our DEP experiments.


Figure 2
View larger version (43K):
[in this window]
[in a new window]

  		Figure 2. Cell populations, viability, and flow rate in the channel. (A): Cells isolated from the embryonic day 12.5 mouse cortex were maintained as NSPCs or differentiated to form neurons or astrocytes and then immunostained with the following markers: Nestin (green, NSPCs), Map2 (red, neurons), or glial fibrillary acidic protein (green, astrocytes). All cell nuclei were stained with Hoechst (blue). (B): Dissociated cells (NSPCs, neurons, and astrocytes) were incubated in dielectrophoresis (DEP) buffer at room temperature for the times indicated, and cell viability was monitored by trypan blue exclusion. Viable cells were expressed as a percentage of the total cells. (C): Mouse NSPCs were trapped at three separate frequencies (1, 5, and 10 MHz) and at flow rates ranging from 0.1 to 10 µl/minute. Only at the highest flow rate (10 µl/minute) and lowest frequency (1 MHz) was the hydrodynamic force sufficient to disrupt the DEP trapping force. (D): A fluid flow simulation program was used to model the fluid dynamics in the channel (described in Materials and Methods). The top panels show the fluid flow in a cross-section of the channel (fluid flow is in the z-direction). The lower panel (graph) depicts the shear stress at the bottom of the channel at flow rates ranging from 0.1 to 10 µl/minute. Abbreviations: cm, centimeter; dyn, dynes; Max, maximum; Min, minimum; min, minute; NSPC, neural stem/precursor cell; sec, second.

 
Cells were introduced into and moved through the DEP channel by fluid flow, and we initially regulated fluid flow manually. However, more reliable results were obtained by regulating the flow rate via a syringe pump attached to the channel outlet. We tested whether fluid flow in the channel would interfere with cell trapping along the electrodes by trapping cells with high-frequency DEP forces and inducing fluid flow at various velocities. Flow rates between 0.5 and 5 µl/minute at all three frequencies tested did not disrupt the trapping of cells (Fig. 2C). At the highest flow rate (10 µl/minute), there was a slight disruption of cell trapping at frequencies of 5 and 10 MHz but a significant decrease in the number of cells trapped at 1 MHz. Cells trapped by DEP forces along the electrode arrays in the channel will experience a shear stress from the fluid flow. Since high levels of shear stress can damage cells, we performed a simulation to determine the hydrodynamic forces generated at the electrodes at the bottom of the channel (Fig. 2D). These calculations showed that the fluid shear stress experienced by cells trapped at the electrodes when the flow rate is 2 µl/minute is 3.05 dyn/cm2 (0.3 Pa), which is less than the lowest shear stresses shown to damage NSPCs grown in fluid bioreactors for 5 days (shear stresses of 10 dyn/cm2 and higher decreased proliferation) [31, 32]. Based on these and subsequent analyses, we determined that 2 µl/minute was an optimal flow rate for our DEP device. Since cells are trapped along the electrodes in our DEP device and then released, we tested whether cells would adhere along the electrodes without applied DEP force. Mouse NSPCs in DEP buffer were unable to attach to the channel surface in the DEP device, whether the surface was glass or SU8 (data not shown). Cells released from the DEP trapping force resumed flow along the channel, could be recovered from the channel outlet, and showed no appreciable loss of viability (supplemental online Table 1), which is consistent with previous studies showing that high percentages of viable cells can be recovered and cultured further after DEP [9].

NSPCs, Neurons, and Astrocytes Have Unique Dielectric Properties Detectable by DEP
To test the DEP responses of mouse NSPCs, neurons, and astrocytes in our devices, we measured the number of cells that were trapped by positive DEP at frequencies ranging from 25 kHz to 10 MHz. A representative video showing trapping of NSPCs at a frequency of 200 kHz is available as part of the supplemental online data (supplemental online Video 1). By graphing the percentage of trapped cells as a function of frequency, we could generate trapping efficiency curves that reflect the positive DEP force felt by the cell population. Analysis of the percentage of viable cells showed that the cells that do not trap at high frequencies (5 and 10 MHz) are dead (data not shown) and thus unable to trap, which confirms the ability of DEP to distinguish live versus dead cells [33]. Therefore, we normalized the data for each cell type such that the maximal percentage of trapped cells was set to 100% (non-normalized data are available in supplemental online Fig. 1A).

At high frequencies (5 and 10 MHz), the vast majority of NSPCs, neurons, and astrocytes experienced positive DEP and were trapped at the electrodes (Fig. 3A). However, at lower frequencies (from 25 kHz to 1 MHz) the amount of positive DEP was significantly different among the three cell types (Fig. 3). Astrocytes trapped at the lowest frequencies (≥95% trapped at 300 kHz), NSPCs trapped at mid-range frequencies (≥95% trapped at 1,000 kHz), and neurons trapped at the highest frequencies (≥95% trapped at 5,000 kHz), thus showing that DEP can be used to distinguish NSPCs from differentiated cells (Fig. 3B). Mouse NSPCs continue to divide in culture, and the data presented in Figure 3 are from cells that had been passaged and in culture for various lengths of time. An analysis of NSPCs by passage number showed that the cells maintained a similar response to DEP over the relatively low passage numbers and time in culture tested here (several weeks; supplemental online Fig. 1B).


Figure 3
View larger version (28K):
[in this window]
[in a new window]

  		Figure 3. Dielectrophoresis trapping efficiency curves distinguish embryonic day 12.5 mouse NSPCs, neurons, and astrocytes. (A): NSPCs, neurons, and astrocytes showed distinct trapping efficiency curves. The clearest differences in the three cell types were evident at the lower frequency range (from 0 to 1,000 kHz). The x-axis of the graph on the left is log scale, whereas that on the right is linear. Curves represent the average from at least three independent experiments, and error bars are SEM. In most cases, all three cell types were isolated from the same set of animals. The table reports significance values between cell types for individual frequencies (**, p < .01; *, p < .05). (B): The approximate frequency at which various percentages (25%, 50%, 75%, and ≥95%) of each of the cell types were trapped is listed. Astrocytes reached ≥95% cells trapped at 300 kHz, whereas NSPCs were at 1,000 kHz and neurons at 5,000 kHz. Abbreviation: NSPC, neural stem/precursor cell.

 
Dielectric Properties of NSPCs from Different Developmental Ages Predict Their Fate Bias
NSPCs at different stages of embryonic development have unique fate potentials. Early cortical NSPCs generate primarily neurons, whereas later in development they give rise predominantly to glia [1–6]. We compared NSPCs from earlier (E12.5) and later (E16.5) developmental time points to determine whether differences in the populations could be detected by DEP. Initial experiments confirmed that the viability of E16.5 NSPCs in the DEP buffer was similar to that of E12.5 NSPCs (data not shown; Fig. 2B). As with the previous cells tested, almost all the E16.5 NSPCs were trapped at higher frequencies (>1 MHz). However, in the 0–1,000 kHz range, the E16.5 NSPCs exhibited positive DEP at lower frequencies than the E12.5 NSPCs (Fig. 4A). The E16.5 NSPC trapping efficiency curve was similar to that of astrocytes in that most cells were trapped at frequencies below 400 kHz and the slope of the curve was steeper than that of E12.5 NSPCs or neurons. To assess whether any difference could be detected in these cells by commonly used markers, we stained the same E12.5 and E16.5 NSPCs used in DEP experiments with antibodies to Sox2 and Nestin (to detect stem and progenitor cells) and GFAP (for astrocytes and some late-embryonic or adult progenitors) (Fig. 4B). There was no difference between the E12.5 and E16.5 cultures in the expression level or percentage of cells expressing Sox2 or Nestin. Furthermore, no cells that immunostained for GFAP were observed in the E12.5 or E16.5 NSPC cultures, confirming that astrocyte contamination was not responsible for the lower frequency trapping of E16.5 NSPCs.


Figure 4
View larger version (33K):
[in this window]
[in a new window]

  		Figure 4. Mouse NSPCs isolated from different developmental ages showed distinct dielectrophoresis (DEP) profiles. (A): Cells isolated from a later time point (E16.5) trapped at lower DEP frequencies than those from earlier in development (E12.5; the curve is the same as shown in Fig. 3) (*, p < .05). The x-axis of the top graph is log scale, whereas that of the bottom graph is linear. The table lists the approximate frequency at which various percentages (25%, 50%, 75%, and ≥95%) of each of the cell types were trapped. E16.5 NSPCs reached ≥95% cells trapped at 400 kHz, whereas E12.5 NSPCs were at 1,000 kHz. (B): E12.5 or E16.5 NSPCs were immunostained for Sox2, Nestin, or GFAP. The top left grayscale images show fields costained with Hoechst to show all nuclei and Sox2 for NSPC nuclei. The top right panels show Nestin (green) overlaid with nuclei (Hoechst; blue). The lower panels show GFAP (red) with nuclei (Hoechst; blue). Astrocytes served as a positive control for GFAP staining, and since they are very spread, flat cells, their nuclei appear larger than those of NSPCs. Abbreviations: E, embryonic day; GFAP, glial fibrillary acidic protein; NSPC, neural stem/precursor cell.

 
Detection of Heterogeneity in a Cell Population by DEP
Our analysis of the dielectric properties of the different cell types revealed that the trapping efficiency curves of some cells (such as astrocytes) were steeper than those of other cells (such as neurons) (Fig. 3A). Astrocytes have been traditionally viewed as fairly homogeneous cells, whereas neuronal cultures are more heterogeneous, since there are several different types of cortical neurons that differ greatly in their function and protein expression (described in Discussion). We hypothesized that the trapping efficiency curve for a homogeneous population of cells would be steep, because the majority of cells would trap at the same frequency because of their similarities in protein expression, cell size and shape, and membrane characteristics. We tested this hypothesis by measuring the DEP trapping efficiency of a cell line that we expected to be relatively homogeneous (human embryonic kidney 293 cells). We found that the trapping efficiency curve for 293 cells was steep, similar to the astrocyte trapping efficiency curve and in contrast to the neuron curve (Fig. 5A).


Figure 5
View larger version (32K):
[in this window]
[in a new window]

  		Figure 5. Dielectrophoresis (DEP) profiles reflect cellular heterogeneity in the population. (A): The DEP lower frequency trapping efficiency curve for a relatively homogeneous population of cells (293 cells) had a steep slope similar to that of astrocytes and steeper than the neuronal curve (curves for astrocytes and neurons are the same as shown in Fig. 3). (B): A mathematical simulation generated DEP trapping efficiency curves for three hypothetical homogeneous cell populations termed cell 1, cell 2, and cell 3. (C): The mathematical simulation of a mixed cell population containing a ratio of 30:30:40 of cells 1:2:3 generated a trapping efficiency curve with a more gradual slope. (D): The ratios of cells in the mixed cell population were varied, and the trapping efficiency curves of the mixtures were graphed. All ratios are presented as cells 1:2:3. Mixtures in which cell 3 was maintained at 40% are graphed in red, and those at 70% are graphed in blue.

 
We next hypothesized that trapping efficiency curve parameters might be predictive of the heterogeneity of the cell sample being analyzed and devised a mathematical simulation to test the hypothesis (described in Materials and Methods). We generated trapping efficiency curves for three different hypothetical and homogeneous cell populations that experienced maximal trapping at ~400 (cell type 1), 600 (cell type 2), or 800 kHz (cell type 3) (Fig. 5B). We next simulated the trapping efficiency curve that would result from a mixture of 30% cell type 1, 30% cell type 2, and 40% cell type 3 and determined that the heterogeneous mixture of cells would have a more gradual trapping efficiency curve slope (Fig. 5C). To test the sensitivity of trapping efficiency curves to changes in the amounts of the different cell populations, we altered the ratio of the hypothetical cells in the simulation and found that changes of 5% in the relative amounts of the cells would cause a detectable shift in the curves (Fig. 5D). We tested the validity of the mathematical model by comparing simulated and experimental mixtures of cells and found that the simulated curve correlated well with the experimental results (supplemental online Fig. 2).


   DISCUSSION
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
Disclosure of Potential...
 Acknowledgments
 References
 
In this study, we assessed the dielectric properties of stem cells of the neural lineage and their differentiated progeny and were able to detect differences among these cells. DEP profiling showed that NSPCs efficiently trapped at mid-range frequencies (~100% trapped at 1,000 kHz) that distinguished them from their differentiated progeny (neurons trapped at higher frequencies [~100% trapped at 5,000 kHz], and astrocytes trapped at lower frequencies [~100% trapped at 300 kHz]). Prasad et al. used a different type of DEP device and studied differentiated cells isolated from postnatal rat but found, similarly, that astrocytes experienced positive DEP at lower frequencies than neurons [14]. Considering the differences in the cells, devices, DEP buffers, and methodology for determining DEP attraction, the frequencies they report for positive DEP for their cells are surprisingly similar to ours (152 kHz for astrocytes, 4,600 kHz for neurons; compare with Fig. 3B). Similar to astrocytes and neurons, glioma cells experienced positive DEP at lower frequencies, whereas neuroblastoma cells required higher frequencies (Fig. 1C) [34]. Notably, our differentiated mouse neurons and mouse neuroblastoma cells experienced positive DEP at similar frequencies (~1,000 and ~900 kHz, respectively) even though the neuroblastoma cells are proliferative and the neurons are not. These data suggest that the cell biological differences between astrocytes and neurons that are detected by DEP are maintained in the cancerous state.

The NSPCs isolated from different developmental ages (E12.5 and E16.5) showed no differences in Nestin and Sox2 expression (Fig. 4) and had similar growth characteristics in vitro (data not shown). However, during embryogenesis, cortical E12.5 NSPCs are more likely to form neurons, and E16.5 NSPCs preferentially form astrocytes, suggesting that phenotypically unique lineage-biased progenitors are present at these stages [1–6]. Remarkably, the dielectric properties of E12.5 and E16.5 NSPCs were distinct and followed the trend of the cells they will preferentially differentiate into (E16.5 NSPCs trap at lower frequencies, much like astrocytes, and E12.5 NSPCs trap at higher frequencies, as do neurons). GFAP-positive cells were not detected in either the E12.5 or E16.5 NSPCs, ruling out the presence of contaminating astrocytes in the E16.5 cultures that might shift the trapping efficiency curve to lower frequencies. Previous studies provide evidence that NSPC changes during development can be related to the differentiated cell types they produce. NSPCs in vivo induce GFAP expression by E15.5 (at the beginning of gliogenesis), based on ganciclovir effects in GFAP promoter-thymidine kinase transgenic mice [26], although we did not detect GFAP expression immunocytochemically in our E16.5 cultures (Fig. 4). Our results suggest that DEP can detect changes in cell phenotype that precede the accumulation of sufficient levels of GFAP to be detected by immunostaining. These data also highlight the fact that stem cell differentiation is a gradual process and that cells may begin to develop aspects of their more differentiated progeny before some markers can be detected and well before they become fully differentiated.

Although some progress has been made toward identification or isolation of glial- or neuron-restricted progenitors, useful markers to distinguish these cells are limited. Several studies have suggested that glial-restricted progenitors (GRPs) are A2B5-positive, tend to be Nestin-positive and GFAP-negative, and preferentially differentiate into glia [35, 36]. However, some A2B5-expressing cells can generate neurons [37], and some cells positive for the astrocytic marker GFAP in late embryonic stages and in the adult are able to produce both neurons and glia. Cells that preferentially generate neurons were shown to express polysialic acid-neural cell adhesion molecule (PSA-NCAM, also termed E-NCAM) [35, 38], but these cells appear to have limited proliferative potential [39], and this marker is unable to distinguish neuron-restricted progenitors (NRPs) from differentiated neurons [35]. Recent work suggests that coexpression of the stem cell markers CD133 or LeX (SSEA1/CD15) in combination with high levels of the neuronal marker CD24 may enrich embryonic neuronal progenitors [39] but is unlikely to identify NRPs at all stages, since adult subventricular zone cells able to generate neurons are CD133-negative [40]. Furthermore, CD133 and CD24 are also expressed on glioblastoma and other tumorigenic cells [39, 40]. Although many of these markers may prove useful for the clear identification of GRPs and NRPs when combined with the presence and absence of additional markers, the neural stem cell field is clearly lacking the necessary numbers of useful markers such as are used for hematopoietic stem cell lineages, where multiple markers are assessed to assign a particular lineage. The problem becomes even more complex and critical when human embryonic stem cells are differentiated into NSPCs that could be used for therapeutic purposes [41]. Identification of lineage-restricted progenitors by an alternative method, such as DEP, may help to provide further characterization of these cells and a means to identify more complete panels of unique markers.

Our data suggest that DEP can be used to monitor and measure cell heterogeneity in stem cell cultures. The trapping efficiency curves for astrocyte cultures (Fig. 4) and 293 cells (Fig. 5) are fairly steep, whereas the curves of neurons and NSPCs have a more gradual slope. Despite some differences [42, 43], astrocytes are relatively homogeneous compared with NSPCs and neurons in culture. NSPC cultures typically contain multipotent stem cells, more committed progenitors, and a subset of fully differentiated cells [7], and the shape of the NSPC trapping efficiency curve measured here is consistent with the diversity of cell types within neurospheres and among the progenitor cells that make up the majority of neurosphere cells. A variety of neuronal subtypes that differ greatly in function, protein expression, size, and morphology are present in cortical neuron cultures. For example, cultures from E12.5 cortex such as those studied here have been shown to contain Cajal-Retzius layer 1 neurons, upper layer 2/3 neurons, layer 5 pyramidal neurons, and layer 6 neurons [44]. Our mathematical simulation (Fig. 5) shows that mixing pure populations of cells (with steep trapping efficiency curves) to increase heterogeneity would result in a summed, less-steep trapping efficiency curve (such as that measured for neurons; Fig. 3A). Therefore, DEP may also provide a means to rapidly analyze the heterogeneity of a cell population or measure heterogeneity in cell populations in which cellular subpopulations cannot be revealed by traditional antibody labeling.

Using dielectric properties to identify stem and differentiated cell populations has certain advantages over flow cytometry and FACS. DEP analysis does not require a large number of cells or expensive equipment, can easily exclude dead cells [33], and has a track record of distinguishing populations of cells that have minor differences [10–13, 15]. DEP provides a rapid, nonbiased screening method that depends on the cell's overall characteristics and does not require any cell-specific probes, such as antibodies. Our findings indicate that DEP may be particularly useful for analysis of stem cell populations that differ from each other but are not distinguished by traditional FACS approaches. Examples of such populations include young and aged long-term reconstituting hematopoietic stem cells that can be isolated by the same panel of five markers using FACS but are functionally distinct because of DNA damage [45]. DEP may also be used to isolate discrete populations of cells to enable development of markers that can be used in traditional approaches, such as FACS and immunostaining. Furthermore, DEP can be combined with flow cytometry or FACS to provide independent and complementary means to identify or isolate cells. DEP devices such as the one described here can be modified to optimize the balance of DEP force and flow rate for different types of cells. DEP is also tunable, since the generation of a dipole in a cell at a given frequency depends on the conductivity of the surrounding medium (equation 3 available in supplemental online data), and DEP media ranging in conductivity at least from 10 to 1,200 µS/cm have been used to detect differences in cell populations [17, 34].

In summary, we have shown that inherent dielectric properties revealed by DEP can distinguish stem cells and their differentiated progeny and can also identify committed progenitors with different fate potentials. DEP provides a unique measure of cellular heterogeneity in a population and is not reliant on the availability of specific cell-type markers. We suspect that these features of DEP will make it a useful application for any set of stem cells in which cellular subtypes are suspected but not able to be confirmed with available markers and techniques.


   DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Disclosure of Potential...
Acknowledgments
 References
 
The authors indicate no potential conflicts of interest.


   ACKNOWLEDGMENTS
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Disclosure of Potential...
 Acknowledgments
References
 
Support was contributed by the Roman Reed Spinal Cord Injury Research Fund of California. We also acknowledge Roger Shih for assistance with video recording and Gregory Lull for assistance with quantitation. DEP devices were fabricated in the University of California Irvine Integrated Nanosystem Research Facilities. L.A.F. and J.L. contributed equally to this work.


   REFERENCES
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Disclosure of Potential...
 Acknowledgments
 References
 

  1. Walsh C, Cepko CL. Widespread dispersion of neuronal clones across functional regions of the cerebral cortex. Science 1992;255:434–440.[Abstract/Free Full Text]

  2. Luskin MB, Parnavelas JG, Barfield JA. Neurons, astrocytes, and oligodendrocytes of the rat cerebral cortex originate from separate progenitor cells: An ultrastructural analysis of clonally related cells. J Neurosci 1993;13:1730–1750.[Abstract]

  3. Grove EA, Williams BP, Li DQ et al. Multiple restricted lineages in the embryonic rat cerebral cortex. Development 1993;117:553–561.[Abstract]

  4. Levison SW, Chuang C, Abramson BJ et al. The migrational patterns and developmental fates of glial precursors in the rat subventricular zone are temporally regulated. Development 1993;119:611–622.[Abstract]

  5. Qian X, Goderie SK, Shen Q et al. Intrinsic programs of patterned cell lineages in isolated vertebrate CNS ventricular zone cells. Development 1998;125:3143–3152.[Abstract]

  6. Qian X, Shen Q, Goderie SK et al. Timing of CNS cell generation: A programmed sequence of neuron and glial cell production from isolated murine cortical stem cells. Neuron 2000;28:69–80.[CrossRef][Medline]

  7. Campos LS, Leone DP, Relvas JB et al. Beta1 integrins activate a MAPK signalling pathway in neural stem cells that contributes to their maintenance. Development 2004;131:3433–3444.[Abstract/Free Full Text]

  8. Pohl H. Dielectrophoresis.Cambridge, U.K.: Cambridge University Press,1978;.

  9. Pethig R, Markx GH. Applications of dielectrophoresis in biotechnology. Trends Biotechnol 1997;15:426–432.[CrossRef][Medline]

  10. Becker FF, Wang XB, Huang Y et al. The removal of human leukaemia cells from blood using interdigitated microelectrodes. J Phys D Appl Phys 1994;27:2659–2662.[CrossRef]

  11. Becker FF, Wang XB, Huang Y et al. Separation of human breast cancer cells from blood by differential dielectric affinity. Proc Natl Acad Sci U S A 1995;92:860–864.[Abstract/Free Full Text]

  12. Cristofanilli M, De Gasperis G, Zhang L et al. Automated electrorotation to reveal dielectric variations related to HER-2/neu overexpression in MCF-7 sublines. Clin Cancer Res 2002;8:615–619.[Abstract/Free Full Text]

  13. Pethig R, Bressler V, Carswell-Crumpton C et al. Dielectrophoretic studies of the activation of human T lymphocytes using a newly developed cell profiling system. Electrophoresis 2002;23:2057–2063.[CrossRef][Medline]

  14. Prasad S, Zhang X, Yang M et al. Separation of individual neurons using dielectrophoretic alternative current fields. J Neurosci Methods 2004;135:79–88.[CrossRef][Medline]

  15. Yang J, Huang Y, Wang X et al. Dielectric properties of human leukocyte subpopulations determined by electrorotation as a cell separation criterion. Biophys J 1999;76:3307–3314.[Medline]

  16. Heida T, Rutten WL, Marani E. Experimental investigation on neural cell survival after dielectrophoretic trapping. Arch Physiol Biochem 2002;110:373–382.[CrossRef][Medline]

  17. Talary MS, Mills KI, Hoy T et al. Dielectrophoretic separation and enrichment of CD34+ cell subpopulation from bone marrow and peripheral blood stem cells. Med Biol Eng Comput 1995;33:235–237.[CrossRef][Medline]

  18. Stephens M, Talary MS, Pethig R et al. The dielectrophoresis enrichment of CD34+ cells from peripheral blood stem cell harvests. Bone Marrow Transplant 1996;18:777–782.[Medline]

  19. Flanagan LA, Rebaza LM, Derzic S et al. Regulation of human neural precursor cells by laminin and integrins. J Neurosci Res 2006;83:845–856.[CrossRef][Medline]

  20. Flanagan L, Ju Y, Marg B et al. Neurite branching on deformable substrates. Neuroreport 2002;13:2411–2415.[CrossRef][Medline]

  21. Wang L, Flanagan LA, Lee AP. Side-wall vertical electrodes for lateral field microfluidic applications. J Microelectromech Syst 2007;16:454–461.[CrossRef]

  22. Ho CT, Lin RZ, Chang WY et al. Rapid heterogeneous liver-cell on-chip patterning via the enhanced field-induced dielectrophoresis trap. Lab Chip 2006;6:724–734.[CrossRef][Medline]

  23. Reynolds BA, Tetzlaff W, Weiss S. A multipotent EGF-responsive striatal embryonic progenitor cell produces neurons and astrocytes. J Neurosci 1992;12:4565–4574.[Abstract]

  24. Tropepe V, Sibilia M, Ciruna BG et al. Distinct neural stem cells proliferate in response to EGF and FGF in the developing mouse telencephalon. Dev Biol 1999;208:166–188.[CrossRef][Medline]

  25. Doetsch F, Caille I, Lim DA et al. Subventricular zone astrocytes are neural stem cells in the adult mammalian brain. Cell 1999;97:703–716.[CrossRef][Medline]

  26. Imura T, Kornblum HI, Sofroniew MV. The predominant neural stem cell isolated from postnatal and adult forebrain but not early embryonic forebrain expresses GFAP. J Neurosci 2003;23:2824–2832.[Abstract/Free Full Text]

  27. Imura T, Nakano I, Kornblum HI et al. Phenotypic and functional heterogeneity of GFAP-expressing cells in vitro: Differential expression of LeX/CD15 by GFAP-expressing multipotent neural stem cells and non-neurogenic astrocytes. Glia 2006;53:277–293.[CrossRef][Medline]

  28. Laywell ED, Rakic P, Kukekov VG et al. Identification of a multipotent astrocytic stem cell in the immature and adult mouse brain. Proc Natl Acad Sci U S A 2000;97:13883–13888.[Abstract/Free Full Text]

  29. Malatesta P, Hartfuss E, Gotz M. Isolation of radial glial cells by fluorescent-activated cell sorting reveals a neuronal lineage. Development 2000;127:5253–5263.[Abstract]

  30. Bonaguidi MA, McGuire T, Hu M et al. LIF and BMP signaling generate separate and discrete types of GFAP-expressing cells. Development 2005;132:5503–5514.[Abstract/Free Full Text]

  31. Gilbertson JA, Sen A, Behie LA et al. Scaled-up production of mammalian neural precursor cell aggregates in computer-controlled suspension bioreactors. Biotechnol Bioeng 2006;94:783–792.[CrossRef][Medline]

  32. Sen A, Kallos MS, Behie LA. Expansion of mammalian neural stem cells in bioreactors: Effect of power input and medium viscosity. Brain Res Dev Brain Res 2002;134:103–113.[CrossRef][Medline]

  33. Huang Y, Holzel R, Pethig R et al. Differences in the AC electrodynamics of viable and non-viable yeast cells determined through combined dielectrophoresis and electrorotation studies. Phys Med Biol 1992;37:1499–1517.[CrossRef][Medline]

  34. Huang Y, Joo S, Duhon M et al. Dielectrophoretic cell separation and gene expression profiling on microelectronic chip arrays. Anal Chem 2002;74:3362–3371.[Medline]

  35. Mujtaba T, Piper DR, Kalyani A et al. Lineage-restricted neural precursors can be isolated from both the mouse neural tube and cultured ES cells. Dev Biol 1999;214:113–127.[CrossRef][Medline]

  36. Rao MS, Mayer-Proschel M. Glial-restricted precursors are derived from multipotent neuroepithelial stem cells. Dev Biol 1997;188:48–63.[CrossRef][Medline]

  37. Maric D, Maric I, Chang YH et al. Prospective cell sorting of embryonic rat neural stem cells and neuronal and glial progenitors reveals selective effects of basic fibroblast growth factor and epidermal growth factor on self-renewal and differentiation. J Neurosci 2003;23:240–251.[Abstract/Free Full Text]

  38. Mayer-Proschel M, Kalyani AJ, Mujtaba T et al. Isolation of lineage-restricted neuronal precursors from multipotent neuroepithelial stem cells. Neuron 1997;19:773–785.[CrossRef][Medline]

  39. Panchision DM, Chen HL, Pistollato F et al. Optimized flow cytometric analysis of central nervous system tissue reveals novel functional relationships among cells expressing CD133, CD15, and CD24. STEM CELLS 2007;25:1560–1570.[Abstract/Free Full Text]

  40. Pfenninger CV, Roschupkina T, Hertwig F et al. CD133 is not present on neurogenic astrocytes in the adult subventricular zone, but on embryonic neural stem cells, ependymal cells, and glioblastoma cells. Cancer Res 2007;67:5727–5736.[Abstract/Free Full Text]

  41. Pruszak J, Sonntag KC, Aung MH et al. Markers and methods for cell sorting of human embryonic stem cell-derived neural cell populations. STEM CELLS 2007;25:2257–2268.[Abstract/Free Full Text]

  42. D'Ambrosio R, Wenzel J, Schwartzkroin PA et al. Functional specialization and topographic segregation of hippocampal astrocytes. J Neurosci 1998;18:4425–4438.[Abstract/Free Full Text]

  43. Zhou M, Kimelberg HK. Freshly isolated astrocytes from rat hippocampus show two distinct current patterns and different [K(+)](o) uptake capabilities. J Neurophysiol 2000;84:2746–2757.[Abstract/Free Full Text]

  44. Shen Q, Wang Y, Dimos JT et al. The timing of cortical neurogenesis is encoded within lineages of individual progenitor cells. Nat Neurosci 2006;9:743–751.[CrossRef][Medline]

  45. Rossi DJ, Bryder D, Seita J et al. Deficiencies in DNA damage repair limit the function of haematopoietic stem cells with age. Nature 2007;447:725–729.[CrossRef][Medline]





This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Data
Right arrow All Versions of this Article:
2007-0810v1
26/3/656    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Reprints/Permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Flanagan, L. A.
Right arrow Articles by Monuki, E. S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Flanagan, L. A.
Right arrow Articles by Monuki, E. S.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
STEM CELLS THE ONCOLOGIST CME ALPHAMED PRESS JOURNALS