| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Bioengineering Department, University of Pittsburgh;
b Growth and Development Laboratory, Children's Hospital of Pittsburgh, Department of Orthopaedic Surgery, Department of Molecular Genetics and Biochemistry, University of Pittsburgh;
c Department of Radiation Oncology, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania, USA
Key Words. Muscle-derived stem cell • Cytokines • Proliferation • Division time • Mitotic fraction
Johnny Huard, Ph.D, Children's Hospital of Pittsburgh Growth and Development Lab, 3705 Fifth Ave., 4151 Rangos Research Center, Pittsburgh, Pennsylvania 15213, USA. Telephone: 412-692-7830; Fax: 412-692-7095; e-mail: jhuard{at}pitt.edu
 |
ABSTRACT |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
|
INTRODUCTION |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Cell therapy applications are, however, constrained by both the insufficient number of cells obtained from a biopsy and also the cells' limited proliferative capacity. In vitro expansion of this population is paramount to obtain suitable numbers of cells for engraftment. This has also been the focus of researchers working with hematopoietic stem cells. Extensive efforts to expand these cultures have found the greatest promise with cytokine-induced expansion [13-15].
In this study, six growth factors were tested for their ability to stimulate MDSCs: epidermal growth factor (EGF), basic fibroblast growth factor (FGF-2), insulin-like growth factor-1 (IGF-1), FLT-3 ligand, hepatocyte growth factor (HGF), and stem cell factor (SCF). These growth factors had previously been demonstrated to stimulate proliferation of either myogenic precursor cells or stem cells. FGF-2, IGF-1, and HGF have been observed to stimulate proliferation of myogenic precursor cells [16-18]. EGF, FLT-3 ligand, and SCF have been shown to stimulate proliferation of stem cells in the hematopoietic compartment and central nervous system (CNS) [19-21].
Cell growth characterization is frequently modeled as exponential growth. However, the key assumption is that 100% of the cell population is dividing. The use of nonexponential and time-lag models may be used to forego this assumption [22,23]. Such models can further provide estimates of several parameters of cell growth kinetics such as: A) fraction of dividing cells; B) rate of commitment to division; C) distribution of the cells in the cell cycle, and D) degree of synchronization in the cell cycle. We therefore used a nonexponential model to examine the underlying mechanisms associated with an increase in cell numbers. The expansion of MDSC can be attributed to at least two factors: A) shortened division time (DT), and/or B) an increase in the number of mitotically active cells (recruitment into G1). This study focused on these two parameters to investigate the expansion of MDSC under the influence of various cytokines.
Using a cell culture system specifically designed for imaging single cells or colonies of cells over long periods of time, we obtained a video record of cell growth. Directly from the images, measures of the total numbers of cells could be made at any time point, and the DTs could be accurately determined. A nonexponential growth model was then used to obtain the fraction of cells that were mitotically active [23]. Few experimental methods exist for quantitating the fraction of dividing cells. Researchers generally have used the growth data with the exponential model to estimate the doubling time, which was often confused as the cell DT. This numerical model eliminates these assumptions by taking into account the observed DT to generate a growth curve that represents the true growth kinetics. Indeed, when the fraction of dividing cells is 1.0, the exponential model is applicable. However, stem cell populations may be expected to have division patterns such that one daughter cell proceeds to a differentiated lineage, and the other daughter cell maintains a mitotically inactive stem cell phenotype. In this case, the fraction of dividing cells would be predicted to be 0.5. One could speculate on a range of division patterns that may exist for a population of stem cells. This model allows for characterization of the division behavior that will ultimately aid in the elucidation of the self-renewal mechanism required for expansion of these cultures. This information may also be useful in understanding which cytokines can recruit quiescent cells in vivo.
To test the effect of various growth factors on the proliferation of the MDSCs, a primary culture of MDSCs and a clone of MDSC, MC13, were examined in the presence of EGF, FGF-2, IGF-1, FLT-3 ligand, HGF, and SCF. The mechanisms by which the growth factors affected stem cell growth were studied; these were namely changes in DT and changes in the dividing fraction. We then compared the response of the primary stem cells to the cytokines with the response of the cultured stem cells. Our results suggest that MDSC can be expanded in vitro with various cytokines, but the mechanism by which these cells become expanded differs from the freshly isolated (MDSC) and long-time cultured (MC13) MDSC.
|
METHODS |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The growth factors examined were human recombinant EGF (100 ng/ml, GIBCO BRL; Rockville, MD; http://www.tmc.tulane.edu/sif/tulgib.htm), human recombinant FGF-2 (100 ng/ml, GIBCO BRL), murine natural IGF-1 (100 ng/ml, GIBCO BRL), human recombinant FLT-3/Flk-2 ligand (25 ng/ml, Sigma; St. Louis, MO; http://www.sigma-aldrich.com), human recombinant HGF (25 ng/ml, Sigma), and mouse recombinant SCF (25 ng/ml, Sigma).
Derivation of the Population Growth Model
Population size at any time, t, depends on two parameters: A) the DT, and B) the mitotic fraction (
). Our model adapts the model of Sherley et al. [23]. In a variation of the derivation described by Sherley et al., population growth may be described by the series of equations:
 |
 |
 |
 |
 |
With this model, the percentage of mitotically active cells, or the dividing fraction, , is determined using both the cell cycle time and the population growth data (N) which are obtained from the same experiment.
Colony Growth
The MDSCs and MC13 cells were plated in 24-well collagen-coated plates at a density of 450 cells/well (2 cm2) in 10% serum medium, as previously described. Growth factors were added at the time of plating and again added with fresh medium 60 hours after plating. Cells were allowed to adhere for 6-12 hours. Using a microscopic imaging system, time-lapsed visible imaging was obtained for individual cells and subsequently for growing colonies [24]. This system uses a biobox incubator mounted to the stage of the microscope, which is linked to a CCD camera (Automated Cell Technologies, Inc.; Pittsburgh, PA; http://www.automatedcell.com). In these experiments, groups of four to six cells were selected for imaging. Coordinate positions of these view fields were recorded by the CytoWorks software program that subsequently controls the time and position of stage movement. Images of each view field were acquired at 10-minute intervals for 4 days. For each cell type and treatment condition, 18 view fields were selected from six wells. Cell population growth was monitored by counting the total number of cells, N, in the view field at 12-hour intervals.
Cell Cycle Duration
From the video images, 100 cells were then selected and tracked. The DT of each cell was determined by direct observation. The time lapsed between cytokinesis was recorded as the length of the cell division cycle.
Mitotic Fraction
To determine the mitotic fraction, , the best fit of the population growth data to the model equation was determined by nonlinear regression using the software package SigmaStat. This program uses the Marquardt-Levenberg algorithm for nonlinear least squares regression analysis.
Approximately six curves for each treatment were fitted using the SigmaStat statistical software package, and the fraction dividing, , parameter was estimated as the average of the six fitted curves.
Statistical Significance Testing
Comparisons of the total cell numbers at each time point were made using the Student's t-test. Cell cycle or DTs were statistically analyzed using the Mann-Whitney rank sum test (p < 0.05 significance level). The dividing fraction was determined using nonlinear regression with the correlation coefficient r2 > 0.90. The Student's t-test was used to determine significant differences for the fractions of dividing cells (p < 0.05 significance level).
|
RESULTS |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
provides a schematic representation of the combinatorial cell culture system and illustrates how data were obtained using this system. The resulting video, for each view field, was a composite of >575 images taken every 10 minutes. From this video, all proliferation data were obtained.
|
shows the equation by which the percentage of mitotically active cells, or the dividing fraction, , is determined using both the cell cycle time and the population growth data (N). According to this model, population size, N, at any time depends on two parameters: A) the DT, and B) the mitotic fraction (). Our combinatorial cell culture and imaging system allow for complete data acquisition of DT and mitotic fraction from any given experiment.
Cell Growth
Primary MDSC
The primary MDSC population size, N, was increased significantly with EGF (72 hours, 96 hours, p < 0.05), and also, to a lesser extent, by FGF-2 (96 hours, p < 0.05), SCF (72 hours, 96 hours, p < 0.05), and IGF-1 (96 hours, p < 0.05) (Fig. 2). Proliferation of the MDSC was significantly inhibited by FLT-3 (96 hours, p < 0.01), while HGF did not significantly affect cell growth of MDSC at any time points tested.
|
Long-Time Cultured MC13
The long-time cultured MC13 cells were also expanded in vitro with FGF-2 (72 hours, 96 hours, p < 0.05), EGF (72 hours, 96 hours, p < 0.05), IGF-1 (72 hours, 96 hours, p < 0.05) and SCF (96 hours, p < 0.05) (Fig. 3). Similar to that observed with MDSC, FLT-3 significantly inhibited the proliferation of MC13 at 96 hours (p < 0.01), while HGF did not significantly inhibit or stimulate cell population growth of MC13 at any time point tested.
|
Cell Cycle Duration
To examine the contribution of DT on proliferation kinetics, cell cycle times in the presence of the growth factors were determined (n = 100). The mean cell DT for nonstimulated MDSCs was determined to be 15.8 ± 3.8 hours; for MC13 cells, the DT was estimated as 16.0 ± 5.3 hours (Fig. 4). The cell cycle times for MDSCs and MC13 cells did not differ significantly (p = 0.43). Of the four growth factors (EGF, FGF-2, IGF-1, and SCF) that expanded the primary MDSC population size, only FGF-2 reduced the DT (13.8 ± 2.5 hours, p < 0.01), suggesting that other mechanisms, such as the recruitment of quiescent stem cells towards proliferation, were responsible for the MDSC expansion (see below). In fact, SCF (which was found capable of expanding the MDSCs) significantly lengthened the DT to the same extent as observed with FLT-3 and HGF, where there was an inhibition of growth and no effect on cell growth, respectively (Fig. 2, Fig. 4). For the long-time cultured MC13 cells, there was a direct correlation between reduction of DT and the expansion of cells. The DT was significantly reduced in the presence of those cytokines which led to an increase of total cell numbers (EGF, FGF-2, IGF-1, and SCF) (Fig. 3, Fig. 4). This result suggests that the expansion of MC13 is primarily made through a decrease of DT.
|
Mitotic Fraction
To examine the contribution of the dividing fraction, , on proliferation kinetics, the effect of growth factors on the percentage of mitotically active cells was also investigated. The total number of cells, N, at 12-hour intervals was obtained as described above. The percentage of dividing/mitotically active cells was then estimated by applying the nonexponential growth model to the growth data. The correlation coefficient, r2, which describes how well the data actually fit to the model such that 0 < r2 < 1.0, was greater than 0.95 in all of the cases with the exception of MDSCs + FGF-2, where r2 > 0.90 (n = 2). The fitted growth curve estimated that 75% of the primary MDSC population was actively cycling. Addition of EGF, IGF-1, or SCF to the medium significantly increased the number of mitotically active MDSC cells to 86% (p < 0.01), 84% (p < 0.04), and 92% (p < 0.001), respectively (Fig. 5A). This likely accounts for the expansion of MDSCs under the influence of these various cytokines, which could not be attributed to DT shortening (Fig. 4A). Interestingly, although HGF stimulation increased the number of dividing cells (82%), the concomitant lengthening of the cell cycle caused a net reduction in population size as compared with nonstimulated cells. The cultured MC13 cell clones had no significant increases in the mitotic fraction, , in the presence of any of the growth factors, suggesting that the observed increases in proliferation are attributed to a reduced cell cycle duration (Fig. 5B).
|
). Our results suggest that the expansion of freshly isolated MDSC is primarily mediated by the recruitment of nonmitotic cells into mitosis while the MC13 are expanded via a reduction in the length of the cell cycle. Consequently, both mechanisms result in an increase in cell number in the MDSC following stimulation with the appropriate cytokines.
|
|
DISCUSSION |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
, suggested that cells in G0 were recruited into G1, whereas a decrease in DT suggested that cells already in the cycle had reduced the length of, or eliminated, a phase of the cycle (e.g., G1 and G2 are commonly absent in cells with DT = 10-12 hours). We have demonstrated here that EGF, FGF-2, IGF-1, and SCF can be used to expand MDSC and MC13 in vitro. Moreover, we have observed the mechanisms by which these growth factors used to expand muscle-derived stem cells (MDSC and MC13) differ; the growth factor decreases the DT in MC13 and increases the mitotic fraction in the MDSC.
Using a novel system to directly observe cells both in culture and in real time, combined with a sensitive mathematical model, we were able to investigate a key parameter of stem cell behavior: recruitment into mitotic activity. The cell culture system that we used here allows for several hundreds of colonies to be imaged during one experiment. Direct observations of divisions are made without the use of calculated estimates based on relative fluorescence, as in BrdU labeling. In this way, the fraction of dividing cells is estimated directly from the same experiment that is used to estimate the cell cycle time. This reduces any errors associated with multiple experimental procedures (e.g., combining BrdU experiments with standard cell/particle counters to establish a growth curve). In addition, the BrdU or [3H] thymidine technique of estimating cell cycle length again assumes that all cells give rise to two daughter cells and uses an exponential model. Importantly, this model uses an additional parameter that provides insight into a mechanism for proliferating cells that is not explained by change in DT alone. Studies examining DT alone could misinterpret the efficacy of using a given growth factor to expand cultures. For example, examining the DT in the presence of SCF may suggest that this growth factor would not be useful in expansion of MDSCs (Fig. 4A). It is revealed here that, although the DT is significantly longer, there is a substantial increase in population size due to the recruitment of cells into mitotic activity (Fig. 5A). Likewise, this model is able to suggest the underlying mechanism of the action of the growth factor, particularly the phase of the cycle in which it exerts its effects.
While we observed an increase in MDSC proliferation in response to SCF, EGF, and IGF-1 by increasing the number of mitotically active cells, the long-term cultured MC13 cells showed no change in the percentage of mitotically active cells in response to these same growth factors. Rather, the MC13 cells responded to these cytokines by shortening their DT (Fig. 6). This suggests that the nondividing fractions differ between the MDSC and MC13 populations. It is conceivable that the nondividing MDSC fraction consists of quiescent stem cells in G0 that may still be recruited into the cell cycle, whereas the MC13 nondividing fraction consists of post-mitotic cells that are terminally committed cells (Fig. 6). This theory would further suggest temporal effects of the growth factors. The growth factors may act at different phases of the cell cycle, depending on the extent of culturing.
The MDSCs are highly stimulated by EGF. Indeed, other groups have demonstrated the ability of EGF to increase proliferation of progenitor stem cells. Takahashi et al. [14] achieved a fivefold increased expansion of hematopoietic progenitors (CD34+) in the presence of EGF. EGF-like repeat motifs (delta-like, dlk) present on stromal cells were observed to promote colony formation by hematopoietic (Sca-1+/linlo/–) stem cells in coculture [25]. Moreover, Reynolds and Weiss [26] have demonstrated that adult CNS stem cells show increased proliferative activity in the presence of EGF. Here, we observed that although EGF stimulation does not reduce the DT of MDSCs, this growth factor can increase the fraction of dividing myogenic stem cells (85% compared with 77% of the control), which ultimately would predict a 10-fold in vitro expansion after 2 weeks. Conversely, the MC13 cell line responded to EGF by a different mechanism. EGF stimulation shortened the DT (12.5 ± 3.5 hours compared with 16.0 ± 5.3 hours in control); however, this was not accompanied by an increase in the number of actively dividing cells. MC13 cultures were expanded most effectively with FGF-2. Interestingly, studies with neural stem cells have found EGF and FGF-2 to have temporally regulated effects on proliferation [21,27]. Gritti et al. [21] have observed that colonies initially formed with EGF stimulation may be further expanded with FGF-2. This suggests that in vitro expansion of stem cell cultures may require the careful timing of growth factor administration.
SCF-induced expansion of the primary MDSCs is attributed to the increase in the mitotically active fraction. Interestingly, the mean DT of these cells is significantly longer than that of nonstimulated cells. Although our putative stem cells did not initially express c-kit [2,3] receptor for SCF, it may be upregulated in the presence of the ligand. Moreover, the long-term cultured MC13 cells responded to SCF by shortening the length of their cell cycle, again suggesting that the effect of the growth factor is temporally regulated in MDSC populations.
Two of the growth factors that we screened in this study demonstrated either inhibitory or no effects on cell proliferation. FLT-3 ligand and HGF have been previously demonstrated to increase proliferation of hematopoietic stem cells [28-30]. However, these studies have generally used cocktails of cytokines. In particular, FLT-3 used in combination with SCF has been shown to increase the number of colony-forming units (similar to an increase in ) and the size of the colonies [30]. These results may also suggest that the MDSC behave differently than the hematopoietic stem cells.
In summary, this study identifies growth factors that can expand MDSC. We have shown that EGF, FGF-2, SCF, and IGF-1 are valuable cytokines to expand both primary and cultured MDSC populations. In particular, EGF stimulation led to the greatest overall increase in cell number in primary cultures as more cells were recruited into the cell cycle. In addition, long-term cultured stem cells may be effectively stimulated with FGF-2 through a significant reduction in DT. Finally, this study illustrates the significance of examining multiple growth parameters to determine whether a given growth factor can be used for the expansion of MDSC.
|
ACKNOWLEDGMENT |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
|
REFERENCES |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  C. Guo, H. Kh. Haider, W. S.N. Shim, R.-S. Tan, L. Ye, S. Jiang, P. K. Law, P. Wong, and E. K.W. Sim Myoblast-based cardiac repair: xenomyoblast versus allomyoblast transplantation. J. Thorac. Cardiovasc. Surg., November 1, 2007; 134(5): 1332 - 1339. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Koellensperger, D. von Heimburg, M. Markowicz, and N. Pallua Human Serum from Platelet-Poor Plasma for the Culture of Primary Human Preadipocytes Stem Cells, May 1, 2006; 24(5): 1218 - 1225. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. N. Salm, P. E. Burger, S. Coetzee, K. Goto, D. Moscatelli, and E. L. Wilson TGF-{beta} maintains dormancy of prostatic stem cells in the proximal region of ducts J. Cell Biol., July 4, 2005; 170(1): 81 - 90. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B.M. Deasy, R.J. Jankowski, T.R. Payne, B. Cao, J.P. Goff, J.S. Greenberger, and J. Huard Modeling Stem Cell Population Growth: Incorporating Terms for Proliferative Heterogeneity Stem Cells, September 1, 2003; 21(5): 536 - 545. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Rosenthal Prometheus's Vulture and the Stem-Cell Promise N. Engl. J. Med., July 17, 2003; 349(3): 267 - 274. [Full Text] [PDF]
|
|
|
|
 |
|
 R. J. Jankowski, B. M. Deasy, B. Cao, C. Gates, and J. Huard The role of CD34 expression and cellular fusion in the regeneration capacity of myogenic progenitor cells J. Cell Sci., November 15, 2002; 115(22): 4361 - 4374. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Z. Qu-Petersen, B. Deasy, R. Jankowski, M. Ikezawa, J. Cummins, R. Pruchnic, J. Mytinger, B. Cao, C. Gates, A. Wernig, et al. Identification of a novel population of muscle stem cells in mice: potential for muscle regeneration J. Cell Biol., May 28, 2002; 157(5): 851 - 864. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| STEM CELLS | THE ONCOLOGIST | CME | ALPHAMED PRESS JOURNALS |
