For a long time, it was believed that the neurological function of the mammalian central nervous system (CNS) could not be restored once it was damaged, e.g., by injury, stroke, or disease. Finding out how to treat and regenerate the damaged CNS has been a subject of intense study in modern medicine (Korr, 1980; Svendsen and Smith, 1999; Björklund and Lindvall, 2000). Some groups have reported recently that transplantation of fetal brain tissue is effective in improving the neurological function of patients with Parkinson's disease (Svendsen and Smith, 1999; Björklund and Lindvall, 2000; Freed et al., 2001). For fetal transplantation to be effective, however, it is necessary to transplant a large volume of fetal brain tissue; material from approximately 10–15 aborted fetuses must be implanted for a single patient (Björklund, 1993). Although fetal transplantation is a scientifically promising way to treat intractable neurological diseases, the associated problems include a lack of donors, and most importantly, the ethics involved in using large amounts of fetal tissue to treat diseases are complicated and controversial (Svendsen and Smith, 1999; Björklund and Lindvall, 2000). To overcome these problems, another approach using cultured fetal brain cells as a source of transplantation material has been pursued (Svendsen and Smith, 1999; Björklund and Lindvall, 2000).
Recent progress in neurobiology has led to a new understanding of the development and cell-lineages of mammalian brain cells and has demonstrated that neurons and glial cells are derived from common immature precursor cells, called neural stem cells (NSCs) (Weiss et al., 1996; Gage, 2000). NSCs are defined as multipotent cells that have the ability to self-renew, which means they have the capacity to differentiate into the three major phenotypes of the CNS: neurons, astrocytes, and oligodendrocytes (Weiss et al., 1996; Gage, 2000). Since the discovery of NSCs, many researchers have been trying to improve damaged mammalian CNS function using ex vivo-expanded immature neural cells containing NSCs (neural stem/progenitor cells; NSPCs) instead of fetal brain tissue (Svendsen and Smith, 1999; Björklund and Lindvall, 2000).
If transplanted human NSPCs are to be used to treat intractable CNS diseases, it is important to develop techniques for human NSPC proliferation in large volumes ex vivo to obtain enough material for clinical use. At present the most widely used culture technique for expanding NSPCs in vitro is the neurosphere method. With this technique, mammalian NSPCs are selectively expanded in a serum-free defined medium containing epidermal growth factor (EGF), fibroblast growth factor 2 (FGF-2), or both, giving rise to floating spheroid cell aggregates called neurospheres (Reynolds et al., 1992; Vescovi et al., 1993; Reynolds and Weiss, 1996). Cells within neurospheres derive from the mammalian fetal brain, which includes a population with self-renewing ability and multipotency that are mostly immunopositive for selective phenotypic markers, such as anti-Nestin (Lendahl et al., 1990) and anti-Musashi1 (Kaneko et al., 2000; Kanemura et al., 2001) antibodies. Some researchers have already reported the isolation of human NSPCs from fetal and adult brain tissues using this method (Svendsen et al., 1998; Carpenter et al., 1999; Johansson et al., 1999; Quinn et al., 1999; Vescovi et al., 1999a,b; Ostenfeld et al., 2000).
To process human NSPCs for clinical use, it is important to establish how long these cells can proliferate ex vivo in a stable manner without losing their capacity to proliferate and undergo transformation. For this, we need to be able to determine the proliferative condition and multipotency of human NSPCs routinely, as precisely and easily as possible. Furthermore, because these cells are cultured using the neurosphere method, we need a practical way of assessing viable cell numbers within neurospheres, using intact neurospheres.
We have isolated NSPCs from human fetal brain tissue using the neurosphere technique and estimated the proliferative activity of these cells by several methods that indirectly measure the number of cells present within intact neurospheres without breaking the neurospheres and without using a radioactive tracer. To measure cell viability, we carried out two assays that indirectly estimate the number of viable cells based on their metabolic activity: the WST-8 assay, in which the reduction of the water-soluble tetrazolium salt WST-8 produces a formazan dye, which reflects the dehydrogenase activity (Marks et al., 1992; Petty et al., 1995; Holst-Hansen and Brönner, 1998; Isobe et al., 1999) and an ATP assay, which measures the ATP content of the total cell plasma (Crouch et al., 1993; Petty et al., 1995). We also examined the population doubling times of human NSPCs using these assays and compared the results with the proliferative activity estimated from measuring DNA synthesis using the 5-bromo-2′-deoxyuridine (BrdU) incorporation assay. We then used the WST-8 and ATP assays to examine changes in the doubling times of human NSPCs in long-term culture.
The results of the present study indicate that these indirect measurements of viable NSPCs based on their metabolic activities, especially the method that measures the total ATP content of total cell plasma, are very convenient, accurate, and useful techniques to estimate the proliferative condition of these cells in large-scale culture. These techniques are likely to contribute to the clinical use of human NSPCs.
MATERIALS AND METHODS
Approval to use human fetal neural tissues was obtained by the ethical committees of both the Osaka National Hospital and the Tissue Engineering Research Center. Tissue procurement was in accordance with the Declaration of Helsinki and in agreement with the ethical guidelines of the Network of European CNS Transplantation and Restoration (NECTAR) and the Japan Society of Obstetrics and Gynecology.
Forebrain tissues from human fetuses aged 7–10 gestational weeks (GW) were obtained from routine legal terminations carried out in the Osaka National Hospital. The fetal brain tissue samples were dissected mechanically in Dulbecco's modified Eagle medium (DMEM)/Ham's F-12 (1:1). After dissection, the tissue samples were digested enzymatically with 0.05% trypsin/0.53 mM EDTA (Invitrogen, Carlsbad, CA) for 20 min at 37°C. The trypsin activity was stopped by adding soybean trypsin inhibitor (2.8 mg/ml, Roche Diagnostics, Mannheim, Germany). After three washes in DMEM/F-12 (1:1), the tissue samples were triturated using a fire-polished Pasteur pipette, then passed through a 40 μm nylon mesh (Cell Strainer; Becton-Dickinson, Franklin Lakes, NJ) to obtain single-cell suspensions. Cell numbers and viability were assessed by trypan blue dye exclusion using a hemocytometer.
Cell suspensions were grown using the neurosphere technique (Reynolds et al., 1992; Vescovi et al., 1993; Reynolds and Weiss, 1996). The growth medium was a defined DMEM/F-12 (1:1)-based medium supplemented with: human recombinant (hr-) EGF (20 ng/ml; Invitrogen); hr-FGF-2 (20 ng/ml; PeproTech Inc, Rocky Hill, NJ); hr-leukemia inhibitory factor (LIF) (10 ng/ml; Chemicon International, Inc., Temecula, CA); heparin (5 μg/ml; Sigma, St. Louis, MO); B27 supplement (Invitrogen); 15 mM HEPES; penicillin (100 U/ml; Invitrogen); streptomycin (100 μg/ml; Invitrogen); and amphotericin B (250 ng/ml; Invitrogen) (Svendsen et al., 1998; Carpenter et al., 1999; Vescovi et al., 1999a,b). Viable single cells at a density of 2 × 106 cells/ml were seeded into uncoated T75 culture flasks and incubated at 37°C in 5% CO2-95% air. Half of the volume of culture medium was replaced by fresh growth medium every 4–5 days. For passaging, every 14–21 days neurospheres were dissociated to single cells by digestion with 0.05% trypsin and 0.53 mM EDTA (Invitrogen) at 37°C for 20 min, and resuspended in 50% fresh growth medium plus 50% neurosphere-conditioned medium.
In every passaging, aliquots of dissociated NSPCs were stored in serum-free freezing medium containing 10% DMSO (Banbanker™, NIPPON Genetics Co. Ltd., Tokyo, Japan) and frozen in liquid nitrogen. Cryopreserved cells were thawed quickly by immersion in a 37°C water bath, rinsed twice with DMEM/F-12 (1:1), and gently resuspended in 50% fresh growth medium plus 50% neurosphere-conditioned medium.
Cell Viability Assay
Single-cell suspensions were prepared from neurospheres by enzymatic dissociation with trypsin followed by passaging through a 40 μm nylon mesh, as described above. The number of viable cells in the single-cell suspensions was determined by cell counting using trypan blue dye exclusion. The number of viable cells was also measured indirectly by following two different metabolic products. In the WST-8 assay, we measured the amount of formazan dye that was produced in the presence of an electron carrier when a highly water soluble tetrazolium salt, WST-8 (2-(2-methoxy-4-nitrophenyl)-3-(4-nitrophenyl)-5-(2,4- disulfophenyl)-2H-tetrazolium, monosodium salt), was reduced by dehydrogenase (Marks et al., 1992; Petty et al., 1995; Holst-Hansen and Brönner, 1998; Isobe et al., 1999). In the ATP assay, the total ATP content in viable cells was measured (Crouch et al., 1993; Petty et al., 1995).
To make a standard curve for each method, 100 μl of single-cell suspension at a density of 2 × 105 cell /ml was plated in a 96-well microplate, and serial 2×-dilutions (from 2 × 105 to 1.25 × 104cell /ml) were made in the wells. After incubating the microplate at 37°C in 5% CO2-95% air for 24 hr, the WST-8 assay (Cell Counting Kit-SF; Tesque Nacalai, Kyoto, Japan) and the ATP assay (CellTiter-Glo™ Luminescent Cell Viability Assay; Promega, Madison, WI) were carried out according to the manufacturer's instructions. Briefly, for the WST-8 assay, 10 μl of the Cell Counting Kit solution was applied to each well in the assay plate and incubated for 5 hr at 37°C. Absorbance was measured at 450 nm using a microplate reader (Benchmark Microplate Reader; BIO-RAD) with a reference wavelength at 655 nm. For the ATP assay, 100 μl of CellTiter-Glo™ Reagent was added to each well of another assay plate and the plate was incubated for 30 min at room temperature. The luminescence signals were detected using a chemiluminescence detection system (Wallac 1420 ARVOSX; Perkin-Elmer).
For the cellular growth curve, viable cells at three different cell densities (1 × 105, 5 × 104 and 2.5 × 104 cells/ml) were incubated for various times (24–168 hr), then the number of viable cells was determined by both the WST-8 and the ATP assay.
BrdU Incorporation Assay
The BrdU incorporation assay was carried out using a Cell Proliferation ELISA BrdU kit (Roche) according to the manufacturer's instructions. Briefly, to make a standard curve serial 2× dilutions of a single-cell suspension were made in a 96-well plate, as in the cell viability assays. After 24 hr of incubation, the cells were labeled with a 10μM BrdU solution and incubated for an additional 8 hr at 37°C. After centrifugation, the labeling medium was removed from the microplates, the cells were dried and fixed, and the cellular DNA was denatured with FixDenat solution (Roche) for 30 min at room temperature. A mouse anti-BrdU monoclonal antibody conjugated with peroxidase was added to each well and the plates were incubated again at room temperature for 2 hr. After washing the plates, tetramethylbenzidine (TMB) was added and the cells were incubated for 30 min at room temperature. Finally, the total reaction was stopped by adding 10 μl 1M H2SO4 to each well. The absorbance of the samples was measured by a microplate reader (BIO-RAD) at 450 nm (reference wavelength 655 nm).
Determination of Population Doubling Time
The population doubling time (DT) was determined independently by the three methods mentioned above (the WST-8 assay, the ATP assay, and the BrdU incorporation assay). Aliquots of 100 μl of single-cell suspension (5 × 104cell/ml) were plated in each well of a 96-well plate. After incubating the plate for 24 hr (t1), the number of viable cells and BrdU labeled proliferative cells were calculated independently using the three methods (N1). The number of viable cells and BrdU labeled proliferative cells (N2) were estimated again using another plate that had been incubated for 144 hr (t2). The DT was determined by the following formula (Cristofalo et al., 1998).
For the estimation of DT in long-term culture, the DT of cells after various incubation days in vitro (DIV) were determined by the ATP assay and the WST-8 assay.
To induce differentiation, neurospheres were plated on polyornithine-coated glass coverslips and cultured in the same medium, lacking the three growth factors and supplemented with 1% FBS. The medium was changed after 7 days, and the cells were cultured for a total of 14 days before fixation for immunocytochemistry (Svendsen et al., 1998; Carpenter et al., 1999; Vescovi et al., 1999a,b).
Human neurospheres and coverslips with differentiated cells were fixed in PBS containing 4% paraformaldehyde for 20 min at room temperature. After fixation, the neurospheres were dipped in 30% sucrose/PBS at room temperature for 30 min, embedded in optimal cutting temperature (OCT)-compound, cut into 12 μm thick sections on a cryotome, and mounted on glass slides. Frozen sections of neurospheres were incubated overnight with the following three primary antibodies: anti-human Nestin (1:500; rabbit polyclonal; Nakamura et al., submitted), anti-Musashi1 (1:500; rat IgG monoclonal; Kaneko et al., 2000; Kanemura et al., 2001) and TO-PRO-3® iodide (1 μM, Molecular Probes Inc., Eugene, OR) in PBS containing 0.1% Triton X and 10% normal goat serum at 4°C. The coverslips were incubated with another set of three primary antibodies: anti-tubulin βIII (TuJ1) (1:500; mouse IgG monoclonal, BAbCO, Richmond, CA), GFAP (1:80; rabbit polyclonal, Sigma), and TO-PRO-3® under the same conditions. After three washes, the neurosphere sections and coverslips were incubated with the appropriate secondary antibodies (Alexa Fluor® 488 goat anti-mouse IgG, 488 goat anti-rabbit IgG, 568 goat anti-rabbit IgG, and 568 goat anti-rat IgG; Molecular Probes) at room temperature for one hour. Fluorescent signals were detected with a confocal scanning laser microscope (LSM510, Carl Zeiss).
The quantitation of the different phenotypes was accomplished by counting the immunolabeled cells on each coverslip. For each condition, three coverslips were evaluated. Four separate randomly chosen fields were counted on each coverslip using a 20× objective. The number of nuclei was counted using TO-PRO-3® staining and the number of tubulin βIII and GFAP immunoreactive cells in each field was also counted. These numbers were summed to give a total representative count for each coverslip. The percentage of each phenotype was determined and these numbers were used to generate mean values for each condition.
Statistical analyses were carried out using analysis of variance (ANOVA) or Student's t-test.
Isolation of Human NSPCs From Fetal Forebrain Tissues
We were successful in isolating four independent lines of NSPCs from forebrain tissues of fetuses from 7–10 GW. The cell viability for all of the primary cultures was over 90% at the beginning of the experiment. There was no difference in the primary cell viability among the cell lines generated from tissues of different gestational age or intervals from the time of termination to the beginning of culture. By almost 7–14 DIV, neurospheres formed from the single-cell suspensions of all four cell lines (Fig. 1A). Most of the cells in the neurospheres were doubly immunopositive for Nestin and Musashi1 antibodies (Fig. 1B–E). Single-cell suspensions created from dissociated neurospheres after every passaging were able to form new neurospheres in all four cell lines (data not shown). In a differentiation assay, tubulin βIII-positive cells (neurons), GFAP-positive cells (astrocytes), and a few nonlabeled cells, which are suspected to be mostly oligodendrocytes, were generated from neurospheres forming cells, indicating the multipotency of these cell lines (Fig. 5C).
Comparison of Different Methods to Estimate Viable Cell Numbers Derived From Human Neurospheres
Our first use of the standard curves for the WST-8 and ATP assay was to examine whether these indirect measurements of viable cell number based on metabolic activity are accurate for human NSPCs. Using serially diluted cells, we determined that there was a direct relationship between the number of viable cells derived from neurospheres and the absorbance measured by the WST-8 assay. There was a significant linear relationship (r2 = 0.9997) between the 450–655 nm absorbance and the number of viable cells in the range from 0–2 × 106 cells/ml (8 independent cultures, Fig. 2A). There was also a significant relationship (r2 = 0.9974) between the luminescence measured in the ATP assay and the number of viable cells in the same range (8 independent cultures, Fig. 2B).
Estimation of the Proliferative Condition of Human NSPCs Using Indirect Measurement of Viable Cell Numbers
We determined the growth curves for human NSPCs using the WST-8 and the ATP assays, and examined the relationship between the viable cell number and the incubation time, again looking for a correlation between the assay results. We used cells at 108 DIV without cryopreservation and measured indirectly the number of viable cells started at three different initial cell plating densities (2.5 × 104, 5.0 × 104, 1 × 105 cells/ml) after various culture times (24, 48, 96, 120, 144, and 168 hr) using the WST-8 assay. Human NSPCs showed exponential proliferation from 24–168 hr without saturation, regardless of the initial cell density (Fig. 3A). The same exponential proliferation of human NSPCs was observed using the ATP assay with cells at 252 DIV and without cryopreservation (Fig. 3B).
Calculation of Doubling Times of Human NSPCs Using Indirect Measurement of Viable Cell Numbers
To estimate the growth of human NSPCs objectively, we calculated the population doubling times using both the ATP assay and the WST-8 assay. Human NSPCs at 155 DIV derived from a 9 GW fetus were used for this study. Human NSPCs plated at a density of 5 × 104 cells/ml grew well for 5 days in culture. The estimated DTs were 82.2 ± 2.1 hr (mean ± SEM, 6 independent cultures) from the WST-8 assay (Fig. 4A) and 83.3 ± 2.0 hr (mean ± SEM, 6 independent cultures) from the ATP assay (Fig. 4B). There was no significant difference between these two estimated DTs obtained using different independent measurements (Student's t-test).
There was also no significant difference between DTs of cells that had undergone a single cryopreservation step and cells that had not (data not shown).
Comparison of Human NSPC Doubling Times Calculated by the Indirect Cell Measurements Based On Metabolic Activity With DT Estimated by BrdU-Labeling
To confirm whether the DTs determined from the indirect measurement of viable cells based on metabolic activity were identical to those calculated by the number of proliferative human NSPCs, we measured the proliferation of human NSPCs using the BrdU incorporation assay, another indirect method.
In measurements of serially diluted cells, the absorbance derived from BrdU labeled cells had a close positive correlation with the number of viable cells per well (r2 = 0.9697; Fig. 2C). Human NSPCs at 155 DIV derived from a 9 GW fetus, the same cells used above to estimate DT using the WST-8 and ATP assays, were also used to determine DT by the BrdU incorporation assay. Human NSPCs plated at a density of 5 × 104 cells/ml grew well for 5 days in culture with an estimated DT of 78.7 ± 7.4 hr (mean ± SEM, 6 independent cultures; Fig. 4C). There was no significant difference between these three estimated DTs obtained from different independent measurements (ANOVA; Fig. 4).
Change in the Proliferative Activity and Multipotency of Human NSPCs in Long-Term Culture
Finally, we examined the influence of long-term culture on the proliferative activity and multipotency of human NSPCs. These properties were estimated from the DTs and the percentage of tubulin βIII-positive cells (neurons), GFAP-positive cells (astrocytes), and nonlabeled cells (suspected to be mostly oligodendrocytes) at various culture periods, using cells derived from the forebrain of a 9 GW fetus. For the determination of DTs in the early DIV, we used cells that had undergone a single cryopreservation step and been cultured for over 4 weeks after the cell thawing procedure.
In the early DIV from the initial culture day, the DTs of human NSPCs were comparatively longer than those in the later DIV (113 ± 7.36 hr at 51 DIV, 98.3 ± 2.67 hr at 68 DIV; mean ± SEM, 6 independent cultures; Fig. 5A). In the period between 100 –200 DIV, the DTs of human NSPCs were shorter than the DTs obtained for other culture periods (84.9 ± 6.62 hr at 108 DIV, 84.8 ± 1.57 hr at 155 DIV, 83.3 ± 7.39 hr at 182 DIV, mean ± SEM, 6 independent cultures; Fig. 5A). The doubling times, however, of human NSPCs became longer again after 200 DIV (104 ± 3.06 hr at 252 DIV, 126.1 ± 9.02 hr at 282 DIV; mean ± SEM, 6 independent cultures; Fig.5 A). The DTs of other lines of normal human NSPCs showed similar changes according to culture progression (data not shown). There were significant differences among these estimated DTs obtained at different culture intervals (P < 0.0001, ANOVA).
The percentage of tubulin βIII-positive cells was almost 20% in all cultures without any significant difference (ANOVA) at all time points analyzed for 251 days in culture (Fig. 5B,C; three independent experiments). Most of the other cells were GFAP-immunopositive, and less than 1% were unlabeled at all culture intervals (data not shown). There was no significant change in the percentage of these glial cells according to culture interval (ANOVA).
Indirect Measurements of Viable Cells Are Effective for Estimating the Proliferation of Human NSPCs
Traditionally, in vitro cell proliferation is determined directly by counting the cells. This method is very simple and inexpensive, and it yields the absolute number of cells. Neural stem/progenitor cells, however, are grown in floating cell masses called neurospheres; to determine the number of viable cells compacted in a neurosphere by counting the cells, it is necessary to break down the neurospheres enzymatically or mechanically to obtain a complete single-cell suspension (Svendsen et al., 1997; Carpenter et al., 1999; Quinn et al., 1999; Vescovi et al., 1999b; Ostenfeld et al., 2000). This procedure is laborious and sometimes inaccurate, because it can be very difficult to dissociate all the neurospheres completely, especially neurospheres that have been cultured for a long time. Another method, the [3H]-thymidine incorporation assay, is also used often to estimate the proliferative activity of NSPCs (Svendsen et al., 1998; Gritti et al., 1999; Vescovi et al., 1999b). This method has the advantage of enabling an estimation of the proliferation activity of human NSPCs in intact neurospheres without dissection. This radioisotope method is expensive, however, and although it may be useful for small scale laboratory studies, it may be less suitable for large scale cultures, for which it could be more complicated to handle the biohazardous materials.
We used two different methods that measure the number of viable NSPCs indirectly, the WST-8 and the ATP assay. Tetrazolium salts, like MTT (3-[4,5-Dimethylthiazol-2-yl]-2,5- dipheneyltetrazolium bromide) or WST-8, are metabolized by the mitochondrial enzyme NAD-dependent succinate dehydrogenase to form a colored formazan product (Marks et al., 1992; Petty et al., 1995; Caldwell and Svendsen, 1998; Holst-Hansen and Brönner, 1998). The amount of formazan dye generated by the activity of the dehydrogenases in cells is known to be directly proportional to the number of living cells. WST-8 is reduced by dehydrogenase in cells to yield a yellow formazan, which is soluble in the tissue culture medium, and there is a good correlation between the results of this assay and the thymidine incorporation assay (Isobe et al., 1999). Another method, the ATP assay, indicates the number of viable cells in culture through quantification of the total cytoplasmic ATP. This assay uses a thermostable firefly luciferase and luminescent signals reflect the amount of ATP present (Crouch et al., 1993; Petty et al., 1995).
The metabolic activity in different cell lines varies, therefore a standard curve needs to be determined for each cell line to test for linearity and to measure the slope of the curve (Holst-Hansen and Brönner, 1998). We first tried to estimate whether the WST-8 and ATP assays accurately reflected the number of viable human NSPCs. The data indicated that the number of viable NSPCs was indeed directly proportional to these two metabolic reaction products (Fig. 2). The WST-8 and ATP assays generated growth curves for NSPCs showing an exponentially proliferative phase (Fig. 3). Moreover, the estimated DTs obtained using each method were almost identical (Fig. 4). These findings indicate that the WST-8 and the ATP assays could potentially replace the direct cell counting method to determine numbers of viable human NSPCs.
The use of metabolic assays such as these to measure viable cells can give false results. For example, if human NSPCs show hypermetabolic activity without cell proliferation, a small number of proliferating cells could be masked by an overwhelming majority of nonproliferating cells. Therefore, we next compared the DTs measured by the WST-8 and the ATP assays with the DT determined by the BrdU incorporation assay, an indirect measurement of proliferating cells that monitors cellular DNA synthesis. Our results showed that the DT determined from the BrdU incorporation assay was almost identical to the DT estimated by the WST-8 and ATP assays (Fig. 4). All this data indicated that in addition to direct cell counting or the [3H]-thymidine incorporation assay, it is also possible to determine viable cell numbers and to detect the proliferative condition of human NSPCs indirectly using the WST-8 or ATP assay. In particular, the ATP assay was so convenient and sensitive that we believe it will become a popular method for estimating rapidly the numbers of viable NSPCs in large scale cultures (Petty et al., 1995).
In testing these methods for the indirect measurement of human NSPCs based on metabolic activity, we faced some problems. The absorbance indicating the yield of water soluble formazan in the WST-8 assay and the luminescence showing the total amount of cellular ATP in the ATP assay were both rather low when we carried out the assays using the routine reaction times optimized for common human tumor cell lines. In particular, the absorbance in the WST-8 assay was lower than the reference data derived from common human tumor cell lines. This tendency toward lower absorbance in making indirect measurements of NSPCs using tetrazolium salts was also observed in a study of rodent NSPCs using a classical MTT assay (Caldwell and Svendsen, 1998). From our results, the WST-8 assay needs reaction times at least twice as long as those used in assays of human tumor cells to obtain a satisfactory absorbance when applied to the determination of human NSPC numbers (data not shown). We suspect that this lower sensitivity of the measurement of viable human NSPCs using tetrazolium salts simply reflected the smaller cytoplasmic volume of these cells. This difference may also be due to a relatively lower activity of mitochondrial NAD-dependent dehydrogenase in human NSPCs compared to differentiated human cells; this is a fundamental concern about indirect measurements using tetrazolium salts. There are no detailed reports about the energy metabolism in human NSPCs and we are currently researching this issue. Finally, this phenomenon may indicate that human NSPCs have a higher capacity for pumping tetrazolium salts out of the cytoplasm than other differentiated cells. Recently, it was reported that a subpopulation of hematopoietic stem cells, called side population cells (SP cells), have the unique property of releasing many useful dyes like Hoechst 33324 (Zhou et al., 2001; Bunting, 2002). SP cells in the hematopoietic lineage show a high activity of ATP-binding cassette (ABC) transporters, such as multidrug resistance (MDR), P-glycoprotein (P-gy), and the breast cancer resistance protein (BCRP) (Zhou et al., 2001; Bunting, 2002). Furthermore, it has also been reported recently that SP cells exist within the NSPCs population (Hulspas and Quesenberry, 2000; Bunting, 2002). In a trial study using verapamil, MDR activity in human leukemic cell lines did not interfere with the MTT cell viability assay (Marks et al., 1992); however, there is still uncertainty about the mechanism of the dye efflux function in NSPCs, which needs to be resolved.
Change in Human NSPC DTs in Long-Term Culture
Our data indicate that the proliferative activity of human NSPCs in long-term culture varied according to the number of days in culture, as has been shown for rodent cells (Morshead et al., 2002). The DTs for the earlier DIV were relatively long, and became shorter as the days in culture progressed. The DTs were the shortest between 100–200 DIV, but after that they gradually lengthened again. No change in the multipotency of human NSPCs was observed in long-term culture (Fig. 5).
For their effective use, it is very important to determine how long human fetal NSPCs can be expanded ex vivo. Our findings indicate that human NSPCs could be effectively expanded for almost 200 days without a loss of multipotency and proliferative activity. Because the maximum DT for human NSPCs was approximately 4 days and because it remains very constant for almost 100 days, we know that human NSPCs can be doubled at least 20 times in 100 days. If an effective method for synchronizing the doubling of human NSPCs in culture conditions can be developed, we could theoretically obtain at least 220-fold NSPCs. This means that if we started with enough primary cells to treat a single patient, we could obtain enough human NSPCs for transplantation into one million patients by expanding the cells ex vivo for at least 200 DIV, as reported by another group (Svendsen and Smith, 1999). This calculation may be very optimistic, but nevertheless, the possibility of expanding human NSPCs ex vivo will encourage the pursuit of their clinical use. To apply our findings to cell processing for clinical use, we will examine the suitability and functional ability of long-cultured human NSPCs as donor cells in vivo (Svendsen and Smith, 1999).
The reasons for the fluctuation in the DTs of human NSPCs in long-term culture are not clear. A decrease in the proliferative rate of NSPCs in prolonged culture has been reported by several groups (Sommer and Rao, 2002). For example, a restricted growth potential of NSPCs was seen in some species, like rats and humans, but not in mice (Svendsen et al., 1997). Several groups have reported that human NSPCs cannot be maintained for prolonged periods, becoming senescent or failing to respond to proliferative signals within 1 year (Svendsen et al., 1998; Quinn et al., 1999; Palmer et al., 2001). This restricted growth capacity of human NSPCs may relate to the lower telomerase activity and shortening of the telomere in human cultured NSPCs compared to rodent cells cultured under similar conditions (Ostenfeld et al., 2000; Sommer and Rao, 2002). Another possibility is a cell-intrinsic mechanism, the so-called, “intracellular division limiter” (clock), which limits the proliferative lifetime of progenitors cultured in vitro (Temple and Raff, 1986; Durand et al., 1998; van Heyningen et al., 2001; Sommer and Rao, 2002). Our findings indicated that after 300 DIV, the DTs of human NSPCs lengthened, and it became difficult to obtain enough cells to calculate DTs using the ATP assay. In some cell lines we observed all the cells suddenly attach to noncoated dishes and differentiate without any reagents to promote cell differentiation, on the day after normal passaging had been carried out (data not shown). This limited proliferative response occurs for some progenitor cells and is called the “Hayflick limit” (Hayflick, 1968; Sommer and Rao, 2002).
The acceleration of DTs observed from the early to the middle term in culture may be explained by one or more of the following hypotheses: one is that human NSPCs may adjust to their in vitro environment and undergo a change in their responsiveness to the growth factors. Changes in proliferative activity due to culture environment have been reported for rodent NSPCs (Gritti et. al., 1999; van Heyningen et al., 2001; Sommer and Rao, 2002). It is also reported that neurospheres begin to arise in the absence of growth factors after repeated passaging (Morshead et al., 2002). We cannot ignore the fact that the estimated DTs obtained for the human neurospheres in the present study were the representative values for heterogeneous cell aggregates containing neural stem cells and various progenitor cells (Reynolds and Weiss, 1996). The NSPCs in neurospheres are thought to vary widely in proliferative activity and degree of differentiation, which is not the case with cultures of monoclonal tumor cell lines, in which all the cells are thought to have the same DT. Therefore, our analyses cannot reveal the exact DT of a specific cell type and will only give the relative DT values of the population of immature neural cells. We suspect that changes in DT in long-term cultures may be due to changes in the kinds of cells in the neurospheres. Stem cells are known to have longer DTs than more differentiated progenitor cells (Yoshikawa, 2000; Sommer and Rao, 2002). In long-term culture, the number of viable progenitor cells increases faster than that of stem cells, and this change may affect the DT of the mass. We are presently investigating the relationship between the degree of heterogeneity of human neurospheres and the cell cycles.
Nevertheless, the actual mechanism underlying the change in human NSPC DT in long-term culture is likely to be more complex than is represented by the above explanations. Successful long-term cultures of human NSPCs for over 1 year have been reported by several groups (Carpenter et al., 1999; Vescovi et al., 1999a), and we have also maintained a few cells for over 1 year without senescence, although their proliferative activity decreased (data not shown). It has been reported that stem cells have additional pathways for regulating proliferation and telomere ends in addition to regulating telomerase levels (Goyns and Lavery, 2000; Sommer and Rao, 2002). Moreover, the action of the intracellular division clock is not always irreversible; it can be reversed using specific manipulations (Kondo and Raff, 2000). Thus, the mechanisms regulating the proliferative activity of human NSPCs are still largely unknown and more vigorous examination of this issue will be needed in the future.
Taking all of our data together, we conclude that indirect measurements of viable cells based on their metabolic activity, especially the ATP assay, are very effective and manually reproducible ways to determine numbers of viable human NSPCs in intact neurospheres. Human NSPCs could be expanded for up to 200 days ex vivo without a notable loss in their ability to proliferate and differentiate. For the clinical use of human NSPCs to be soon realized, we must examine the proliferative activity of human NSPCs in long-term and bulk culture systems in more detail.
We thank Ms.Tomoko Shofuda, Atsuyo Maeda, Ema Hirai, Rika Hi, and Jesmin Tajria for their technical assistance. This study was supported in part by the Three-Dimensional Tissue Module Project (A Millennium Project) from the Ministry of Economy, Trade and Industry and the Health Sciences Research Grants (H13-BS-010) from the Ministry of Health, Labor and Welfare of Japan.
Yonehiro Kanemura and Hideki Mori contributed equally to this work.