Printed in Sweden Copyright 0 1975 by Academic Press, Inc. All rights of reproduction in any form reserved
Experimental Cell Research92 (1975) 271-274
POPULATION ANALYSIS OF ARRESTED HUMAN DIPLOID FIBROBLASTS BY FLOW MICROFLUOROMETRY R. T. DELL’ORCO;
H. A. CRISSMAN,2 J. A. STEINKAMP* and P. M. KRAEMERa
lBiomedical Division, The Samuel Roberts Noble Foundation, Inc., Ardmore, Okla. 73401, and 2Los Alamos Scientific Laboratory, University of California, Los Alamos, N.M. 87544, USA
SUMMARY Confluent cultures of human diploid fibroblasts were maintained for 28 days with medium containing 0.5 % serum. Periodically during this time cells were exposed to aH-thymidine for 72 h; harvested; and analysed by flow microfluorometric, ‘cell sorting’, and autoradiographic techniques. The results showed that cells cultured under these conditions maintain a stable population distribution similar to that occurring when a population reaches confluency in medium containing 10% serum. Low labeling indices, sparce grain densities, and the presence of some mitotic cells indicated that a limited amount of cell-cycle traverse did occur but that both the S and G2 phases were prolonged. This new state of reduced mitotic activity with prolonged cell-cycle times may mimic the long-term inhibition of cell-cycle traverse of expanding tissues in vivo.
The mitotic activity of human diploid fibroblasts in culture can be inhibited by reducing the serum concentration of the incubation medium from 10% to 0.1 or 0.5% [l, 21. Cells maintained in this manner for up to 6 months can be induced to re-enter a state of rapid proliferation by subcultivation with medium containing 10 % serum [2, 31. This method for inhibiting division in cultured cells may offer an experimental tool for the investigation of growth regulation in intact animals. This may be true especially for those cells which in vivo exhibit little or no mitotic activity unless exposed to a specific stimulus (i.e., the fibroblasts of connective tissue during wound healing [4]). To develop this systemas an in vitro model, it must be characterized with respect to a number of physiological parameters. Two
such parameters are the completeness of the mitotic block and the phase(s) of the division cycle in which the cells are maintained under experimental conditions. Preliminary experiments employing autoradiographic techniques to determine the number of cells synthesizing DNA, as well as the estimation of mitotic indices [l, 21,indicated that a true non-mitotic state was established by reducing the serum concentration of the incubation medium. The purpose of the experiments described in this report was to confirm the non-mitotic nature of the inhibited population, as well as to determine if the cells were blocked in one particular phase of the division cycle. Analysis of inhibited populations by flow microfluorometry (FMF) revealed that 90 % of the cells had a DNA content consistent with the G 1 phase of the division cycle; howExptl Cell Res 92 (1975)
212
Dell’Orco et al.
ever, this was not a static population. Isolation of the G2+M (mitosis) fraction by ‘cell sorting’ techniques [5] showed that a minimum amount of mitosis did occur during cultivation with the reduced serum concentration but that the time for cell-cycle traverse was prolonged. This new state of J80 x) 40 60 al 40 reduced mitotic activity with prolonged cellcycle times may mimic the long-term inhibi- Fig. 1. Abscissa: DNA content/cell (arb. units); no. of cells. tion of cell-cycle traverse of expanding tissues ordinate: DNA content distribution of CF-1 cultures maintained on medium containing 0.5 % serum for 0, 14, in vivo. MATERIALS AND METHODS
and 28 days. A, Same data as l , except for a IO-fold magnification of the ordinate scales. Crosshatched areas indicate the cells sorted out with a 4C DNA content.
Cellsusedin theseexperimentswerehumandiploid fibroblastsderivedfrom newbornforeskin material of cells in each phase of the division cycle and designatedCF-1. When the cells reachedconfluency, they were routinely subculturedat a 1:4 was comparable to fractions deduced by split ratio in McCoy’sMedium7a [l] supplemented biochemical cell-cycle analysis [7, lo]. The with 10% fetal bovine serum and 50 ,ug/ml of both penicillin. and streptomycin. Mitotic activity was in- major peak represents cells with a DNA hibited in confluent cultures by replacing the medium content consistent with the Gl phase of the containing 10% serum with that containing 0.5 % serum. This medium was replaced twice weekly dur- cell cycle. The peak at twice the mode of the ing the 28 days of the experimental period. The cells G 1 DNA content represents cells in G2 + M. were determined to be free of mycoplasma contaCells in S phase with varying degrees of mination by routine procedures [6]. Each culture was exposed to 0.1 &i/ml of 8H- completion of DNA replication are distribthymidine (New England Nuclear, Boston, Mass.; spec. act. 20 Ci/mmole) for 72 h before harvest. uted between the G 1 and the G2 + M peaks. After the cultures were harvested, the cells were The distribution plots for the three sample rendered monodisperse, stained by the a&lavineFeulgen technique, and submitted to FMF and cell points shown in fig. 1 are representative of sorting by methods previously described [7, 81. Cells the data obtained at all sample points. These collected by the sorting technique were applied to ‘subbed’ microscope slides, and autoradiographs were results showed that exposure to low serum prepared by the method of Puck & Steffen [9] except medium maintained the population distributhat the cells were stained and developed after exposure for 1 week which wasdeterminedto be optimal. The resultant autoradiographs had a very law Table 1. Cell-cycle composition of CF-I cells background. Only those cells with five or more grains were considered to be labeled, and 500 cells were scored for each determination.
maintained on medium containing 0.5 % seruma Phase
RESULTS Fig. 1 shows the DNA distribution patterns obtained by FMF from cells taken at confluency (0 day) and after 14 and 28 days of exposure to medium containing 0.5 % serum. Computer analysis of the DNA content of exponentially distributed cell populations had been used to demonstrate that the ,fraction ExptI Cell Res 92 (1975)
Day
Gl
S
G2+M
i 1:
90.4 92.0 90.7 87.3
Ei 4.8 912
0.5 3.4 4.5 3.5
;:
90.4 89.9
6.5 5.9
:::
o Values are percentages of the observed population detected in each of the designated phases of the cell cycle.
Population analysis of arrested cells
tion that was present when the cells attained confluency in medium containing 10 X serum. When the FMF frequency distribution data for all time points (6, 3, 7, 14, 21, and 28 days) were submitted to computer analysis [lo], the results shown in table 1 were obtained. Throughout the experiment, approx. 90% of the cells had a DNA content of Gl cells (i.e., for the entire 28 days approx. 10 Y0 of the cells at any time were beyond the G l/S boundary). S phase cells, on the basis of DNA content, ranged from 4.6 to 9.2 %. The O-day sample had very few cells with a 4C DNA content (G2 +M fraction), but all other samples gave values ranging from 3.4 to 4.5 %. This population distribution can be contrasted to cells in the log phase of growth which have 40.6 % G 1, 46.9 % S, and 12.5Y0 G2 +M cells. These results indicated that CF-1 cells incubated with medium containing 0.5 % serum maintained a fairly stable distribution of cells among the different phases of the cell cycle for as long as 28 days. This population distribution was similar to that achieved at confluency and was in agreement with Tobey & Ley [l l] who found that in stationary phase culture of CHO cells, more than 90 % of the population was in the G 1 phase of the division cycle. To determine if the population distributions obtained by FMF were those of static populations, we labeled the cells with 3Hthymidine for 72 h before harvest. The labeling indices, as determined by autoradiography for the entire population and for a ‘sorted’ population [5] consisting of cells with a 4C DNA content (G2 +M fraction), are shown in table 2. Despite the FMF evidence that 10% of the cells had a DNA content of greater than 2C, the over-all labeling indices after 72 h of exposure to 3H-thymidine were only 2.6 to 6.0%. With respect to the sorted population of cells with a 4C DNA content, it was determined that less than 1% of these
273
Table 2. Labeling index of CF-I cells exposed to aiY-ihymid~ for 72 h before harvest Day
i
Total population’
ii
G2 + M fraction*
:-;:
7
2:6
1:s
;; 28
4.5 2:
1.8 4.8
a Values are percentages of labeled cells in the total population. Values are the percent of labeled cells in the ‘sorted G2 + M fraction.
cells were in mitosis; therefore, the sorted fraction was composed of essentially all G2 phase cells. The labeling indices from this fraction ranged from 1.4 to 4.8 % during the 28-day experimental period. These data showed that a minimum of 95 ‘X0of the cells in G2 at the time of harvest crossed the S/G2 boundary before the 72-h labeling period. The labeling indices for both the total population and the G2 fraction indicated that cells exposedto medium containing 0.5 % serum were not static populations but did have a minimum number of cycling cells. These results also showed that the cycling cells were traversing the cell cycle very slowly. DISCUSSION The results from the experiments described herein show that human diploid fibroblasts maintained with medium containing 0.5 % serum achieve a stable population distribution for as long as 28 days in culture. This distribution is similar to that occurring when a population reaches confluency under stationary culture conditions; however, this is not a static population. In contrast to preliminary reports which employed the culture conditions described here [l-3] as well as the Exptl Cell Res 92 (1973
274 Dell’Orco et al. work of Westermark with a glia cell system [12], the arrested populations continue to show evidence of slow cell-cycle traverse during the entire experimental period. The low labeling indices, especially those of the G2 sorted cells, the sparse grain density of the cells that were labeled, and the presence of some mitotic cells indicate that a very limited amount of cell-cycle traverse does occur but that both the S and the G2 phases are prolonged. In addition to this, it seems likely that the length of time cells spend in the Gl phase increases also because of the number of cells detected in this fraction. These experiments do not distinguish between the possibilities of a subpopulation moving through the division cycle with the majority blocked in one phase (Gl) or the entire population traversing. However, what is clear is that the cells involved in traverse are proceeding through Gl, S, and G2 at a much reduced rate. The cells of many tissues in vivo are described as being expanding populations. Scattered mitotic cells can be detected in these tissues which have an average daily mitotic rate of less than 0.5 % [13]. It has been postulated that the majority of cells in these tissues are inhibited from cell-cycle traverse at particular phases of the division cycle. These arrested cells have been considered to be in abnormally long Gl or G2 periods, which can be dramatically shortened by the stimulus to proliferate [14]. They have also been envisioned to be in a GO phase which, while distinct from G 1, has certain properties such as DNA content in common with cells in G 1 [15]. In a similar manner, cells in culture achieving confluency and in a stationary
Exptl Cell Res 92 (197.5)
culture condition have been shown to be inhibited from cell-cycle traverse to the same extent and at the same stages of the division cycle [l 1, 16-181. The system employed in our study allows for the long-term maintenance of the population distribution found at confluency in human diploid fibroblast cultures. Since this distribution can be related to that found in certain tissues in situ, it may offer an in vitro model system for the study of the regulatory processesoccurring in these tissues. This work was performed in part under the auspices of the US AEC.
REFERENCES Kruse, P F Jr, Whittle, W & Miedema, E, J cell bio142 (1969) 113. Dell’Orco, R T, Mertens, J G & Kruse, P F Jr, Exptl cell res 77 (1973) 356. - Exptl cell res 84 (1974) 363. Bullough, W S & Laurence, E B, Exptl cell res 21 (1960) 394. Steinkamp, J A, Fulwyler, M J, Coulter, J R, Hiebert, R D, Homey, J L & Mullaney, P F, Rev sci instr 44 (1973) 1301. Hayflick, L, Texas rep biol med 23 (1965) 385. 76:Kraemer, P M, Deaven, L L, Crissman, H A, Steinkamp, J A & Petersen, D F, Cold Spring Harbor symp quant biol 38 (1973) 133. a. Kraemer, P M, Deaven, L L, Crissman, H A & Van Dilla, M A, Adv cell mol biol 2 (1972) 47. Puck, T T & Steffen, J, Biophys j 3 (1963) 379. 1;: Dean, P N & Jett, J H, J cell biol 60 (1974) 523. 11. Tobey, R A & Ley, K D, J cell bio146 (1970) 151. 12. Westermark, B, Exptl cell res 82 (1973) 341. 13. Leblond, C P, Nat1 cancer inst monograph 14 (1964) 119. Gelfant, S & Smith, J G, Science 178 (1972) 357. ::: Pott, H M & Quastler, H, Physiol rev 43 (1963) 357. 16. Macieira-Coelho, A, Proc sot exptl biol med 125 (1967) 548. 17. Temin, H M, J cellular physiol 78 (1971) 161. 18. Rovera, G & Baserga, R, Exptl cell res 78 (1973) 118. Received October 6, 1974 Revised version received November 5, 1974
Experimental Cell Research92 (1975) 271-274
POPULATION ANALYSIS OF ARRESTED HUMAN DIPLOID FIBROBLASTS BY FLOW MICROFLUOROMETRY R. T. DELL’ORCO;
H. A. CRISSMAN,2 J. A. STEINKAMP* and P. M. KRAEMERa
lBiomedical Division, The Samuel Roberts Noble Foundation, Inc., Ardmore, Okla. 73401, and 2Los Alamos Scientific Laboratory, University of California, Los Alamos, N.M. 87544, USA
SUMMARY Confluent cultures of human diploid fibroblasts were maintained for 28 days with medium containing 0.5 % serum. Periodically during this time cells were exposed to aH-thymidine for 72 h; harvested; and analysed by flow microfluorometric, ‘cell sorting’, and autoradiographic techniques. The results showed that cells cultured under these conditions maintain a stable population distribution similar to that occurring when a population reaches confluency in medium containing 10% serum. Low labeling indices, sparce grain densities, and the presence of some mitotic cells indicated that a limited amount of cell-cycle traverse did occur but that both the S and G2 phases were prolonged. This new state of reduced mitotic activity with prolonged cell-cycle times may mimic the long-term inhibition of cell-cycle traverse of expanding tissues in vivo.
The mitotic activity of human diploid fibroblasts in culture can be inhibited by reducing the serum concentration of the incubation medium from 10% to 0.1 or 0.5% [l, 21. Cells maintained in this manner for up to 6 months can be induced to re-enter a state of rapid proliferation by subcultivation with medium containing 10 % serum [2, 31. This method for inhibiting division in cultured cells may offer an experimental tool for the investigation of growth regulation in intact animals. This may be true especially for those cells which in vivo exhibit little or no mitotic activity unless exposed to a specific stimulus (i.e., the fibroblasts of connective tissue during wound healing [4]). To develop this systemas an in vitro model, it must be characterized with respect to a number of physiological parameters. Two
such parameters are the completeness of the mitotic block and the phase(s) of the division cycle in which the cells are maintained under experimental conditions. Preliminary experiments employing autoradiographic techniques to determine the number of cells synthesizing DNA, as well as the estimation of mitotic indices [l, 21,indicated that a true non-mitotic state was established by reducing the serum concentration of the incubation medium. The purpose of the experiments described in this report was to confirm the non-mitotic nature of the inhibited population, as well as to determine if the cells were blocked in one particular phase of the division cycle. Analysis of inhibited populations by flow microfluorometry (FMF) revealed that 90 % of the cells had a DNA content consistent with the G 1 phase of the division cycle; howExptl Cell Res 92 (1975)
212
Dell’Orco et al.
ever, this was not a static population. Isolation of the G2+M (mitosis) fraction by ‘cell sorting’ techniques [5] showed that a minimum amount of mitosis did occur during cultivation with the reduced serum concentration but that the time for cell-cycle traverse was prolonged. This new state of J80 x) 40 60 al 40 reduced mitotic activity with prolonged cellcycle times may mimic the long-term inhibi- Fig. 1. Abscissa: DNA content/cell (arb. units); no. of cells. tion of cell-cycle traverse of expanding tissues ordinate: DNA content distribution of CF-1 cultures maintained on medium containing 0.5 % serum for 0, 14, in vivo. MATERIALS AND METHODS
and 28 days. A, Same data as l , except for a IO-fold magnification of the ordinate scales. Crosshatched areas indicate the cells sorted out with a 4C DNA content.
Cellsusedin theseexperimentswerehumandiploid fibroblastsderivedfrom newbornforeskin material of cells in each phase of the division cycle and designatedCF-1. When the cells reachedconfluency, they were routinely subculturedat a 1:4 was comparable to fractions deduced by split ratio in McCoy’sMedium7a [l] supplemented biochemical cell-cycle analysis [7, lo]. The with 10% fetal bovine serum and 50 ,ug/ml of both penicillin. and streptomycin. Mitotic activity was in- major peak represents cells with a DNA hibited in confluent cultures by replacing the medium content consistent with the Gl phase of the containing 10% serum with that containing 0.5 % serum. This medium was replaced twice weekly dur- cell cycle. The peak at twice the mode of the ing the 28 days of the experimental period. The cells G 1 DNA content represents cells in G2 + M. were determined to be free of mycoplasma contaCells in S phase with varying degrees of mination by routine procedures [6]. Each culture was exposed to 0.1 &i/ml of 8H- completion of DNA replication are distribthymidine (New England Nuclear, Boston, Mass.; spec. act. 20 Ci/mmole) for 72 h before harvest. uted between the G 1 and the G2 + M peaks. After the cultures were harvested, the cells were The distribution plots for the three sample rendered monodisperse, stained by the a&lavineFeulgen technique, and submitted to FMF and cell points shown in fig. 1 are representative of sorting by methods previously described [7, 81. Cells the data obtained at all sample points. These collected by the sorting technique were applied to ‘subbed’ microscope slides, and autoradiographs were results showed that exposure to low serum prepared by the method of Puck & Steffen [9] except medium maintained the population distributhat the cells were stained and developed after exposure for 1 week which wasdeterminedto be optimal. The resultant autoradiographs had a very law Table 1. Cell-cycle composition of CF-I cells background. Only those cells with five or more grains were considered to be labeled, and 500 cells were scored for each determination.
maintained on medium containing 0.5 % seruma Phase
RESULTS Fig. 1 shows the DNA distribution patterns obtained by FMF from cells taken at confluency (0 day) and after 14 and 28 days of exposure to medium containing 0.5 % serum. Computer analysis of the DNA content of exponentially distributed cell populations had been used to demonstrate that the ,fraction ExptI Cell Res 92 (1975)
Day
Gl
S
G2+M
i 1:
90.4 92.0 90.7 87.3
Ei 4.8 912
0.5 3.4 4.5 3.5
;:
90.4 89.9
6.5 5.9
:::
o Values are percentages of the observed population detected in each of the designated phases of the cell cycle.
Population analysis of arrested cells
tion that was present when the cells attained confluency in medium containing 10 X serum. When the FMF frequency distribution data for all time points (6, 3, 7, 14, 21, and 28 days) were submitted to computer analysis [lo], the results shown in table 1 were obtained. Throughout the experiment, approx. 90% of the cells had a DNA content of Gl cells (i.e., for the entire 28 days approx. 10 Y0 of the cells at any time were beyond the G l/S boundary). S phase cells, on the basis of DNA content, ranged from 4.6 to 9.2 %. The O-day sample had very few cells with a 4C DNA content (G2 +M fraction), but all other samples gave values ranging from 3.4 to 4.5 %. This population distribution can be contrasted to cells in the log phase of growth which have 40.6 % G 1, 46.9 % S, and 12.5Y0 G2 +M cells. These results indicated that CF-1 cells incubated with medium containing 0.5 % serum maintained a fairly stable distribution of cells among the different phases of the cell cycle for as long as 28 days. This population distribution was similar to that achieved at confluency and was in agreement with Tobey & Ley [l l] who found that in stationary phase culture of CHO cells, more than 90 % of the population was in the G 1 phase of the division cycle. To determine if the population distributions obtained by FMF were those of static populations, we labeled the cells with 3Hthymidine for 72 h before harvest. The labeling indices, as determined by autoradiography for the entire population and for a ‘sorted’ population [5] consisting of cells with a 4C DNA content (G2 +M fraction), are shown in table 2. Despite the FMF evidence that 10% of the cells had a DNA content of greater than 2C, the over-all labeling indices after 72 h of exposure to 3H-thymidine were only 2.6 to 6.0%. With respect to the sorted population of cells with a 4C DNA content, it was determined that less than 1% of these
273
Table 2. Labeling index of CF-I cells exposed to aiY-ihymid~ for 72 h before harvest Day
i
Total population’
ii
G2 + M fraction*
:-;:
7
2:6
1:s
;; 28
4.5 2:
1.8 4.8
a Values are percentages of labeled cells in the total population. Values are the percent of labeled cells in the ‘sorted G2 + M fraction.
cells were in mitosis; therefore, the sorted fraction was composed of essentially all G2 phase cells. The labeling indices from this fraction ranged from 1.4 to 4.8 % during the 28-day experimental period. These data showed that a minimum of 95 ‘X0of the cells in G2 at the time of harvest crossed the S/G2 boundary before the 72-h labeling period. The labeling indices for both the total population and the G2 fraction indicated that cells exposedto medium containing 0.5 % serum were not static populations but did have a minimum number of cycling cells. These results also showed that the cycling cells were traversing the cell cycle very slowly. DISCUSSION The results from the experiments described herein show that human diploid fibroblasts maintained with medium containing 0.5 % serum achieve a stable population distribution for as long as 28 days in culture. This distribution is similar to that occurring when a population reaches confluency under stationary culture conditions; however, this is not a static population. In contrast to preliminary reports which employed the culture conditions described here [l-3] as well as the Exptl Cell Res 92 (1973
274 Dell’Orco et al. work of Westermark with a glia cell system [12], the arrested populations continue to show evidence of slow cell-cycle traverse during the entire experimental period. The low labeling indices, especially those of the G2 sorted cells, the sparse grain density of the cells that were labeled, and the presence of some mitotic cells indicate that a very limited amount of cell-cycle traverse does occur but that both the S and the G2 phases are prolonged. In addition to this, it seems likely that the length of time cells spend in the Gl phase increases also because of the number of cells detected in this fraction. These experiments do not distinguish between the possibilities of a subpopulation moving through the division cycle with the majority blocked in one phase (Gl) or the entire population traversing. However, what is clear is that the cells involved in traverse are proceeding through Gl, S, and G2 at a much reduced rate. The cells of many tissues in vivo are described as being expanding populations. Scattered mitotic cells can be detected in these tissues which have an average daily mitotic rate of less than 0.5 % [13]. It has been postulated that the majority of cells in these tissues are inhibited from cell-cycle traverse at particular phases of the division cycle. These arrested cells have been considered to be in abnormally long Gl or G2 periods, which can be dramatically shortened by the stimulus to proliferate [14]. They have also been envisioned to be in a GO phase which, while distinct from G 1, has certain properties such as DNA content in common with cells in G 1 [15]. In a similar manner, cells in culture achieving confluency and in a stationary
Exptl Cell Res 92 (197.5)
culture condition have been shown to be inhibited from cell-cycle traverse to the same extent and at the same stages of the division cycle [l 1, 16-181. The system employed in our study allows for the long-term maintenance of the population distribution found at confluency in human diploid fibroblast cultures. Since this distribution can be related to that found in certain tissues in situ, it may offer an in vitro model system for the study of the regulatory processesoccurring in these tissues. This work was performed in part under the auspices of the US AEC.
REFERENCES Kruse, P F Jr, Whittle, W & Miedema, E, J cell bio142 (1969) 113. Dell’Orco, R T, Mertens, J G & Kruse, P F Jr, Exptl cell res 77 (1973) 356. - Exptl cell res 84 (1974) 363. Bullough, W S & Laurence, E B, Exptl cell res 21 (1960) 394. Steinkamp, J A, Fulwyler, M J, Coulter, J R, Hiebert, R D, Homey, J L & Mullaney, P F, Rev sci instr 44 (1973) 1301. Hayflick, L, Texas rep biol med 23 (1965) 385. 76:Kraemer, P M, Deaven, L L, Crissman, H A, Steinkamp, J A & Petersen, D F, Cold Spring Harbor symp quant biol 38 (1973) 133. a. Kraemer, P M, Deaven, L L, Crissman, H A & Van Dilla, M A, Adv cell mol biol 2 (1972) 47. Puck, T T & Steffen, J, Biophys j 3 (1963) 379. 1;: Dean, P N & Jett, J H, J cell biol 60 (1974) 523. 11. Tobey, R A & Ley, K D, J cell bio146 (1970) 151. 12. Westermark, B, Exptl cell res 82 (1973) 341. 13. Leblond, C P, Nat1 cancer inst monograph 14 (1964) 119. Gelfant, S & Smith, J G, Science 178 (1972) 357. ::: Pott, H M & Quastler, H, Physiol rev 43 (1963) 357. 16. Macieira-Coelho, A, Proc sot exptl biol med 125 (1967) 548. 17. Temin, H M, J cellular physiol 78 (1971) 161. 18. Rovera, G & Baserga, R, Exptl cell res 78 (1973) 118. Received October 6, 1974 Revised version received November 5, 1974
