Characterization of the Cell Cycle of Cultured Human Diploid Cells: Effects of Aging and Hydrocortisone GARY L. GROVE AND VINCENT J. CRISTOFALO The Wistar Institute of Anatomy and Biology, 36th Street at Spruce, Philadelphia, Pennsylvania 19104

ABSTRACT Age-related changes in the cytokinetics of human diploid cells in vitro have been compared in normal cultures and in cultures in which lifespan has been prolonged by the addition of hydrocortisone to the medium. For both cultures, with advancing age the fraction of cells in the actively proliferating pool decreased and the intercellular variation in cell cycle times increased. The average cell cycle time was prolonged during aging due almost entirely to changes in the duration of GI. The duration of S remained constant, while a small delay in G, was observed in late passage cells near the end of their lifespan. Although the same pattern of change in proliferative parameters occurred in both control and hydrocortisone-treated cultures, the changes were somewhat delayed in the presence of the steroid. The results are interpreted in terms of several cell cycle models and suggest that the events controlling cell proliferation are sensitive to hydrocortisone modulation during the GI and possibly the G2 periods.

Senescence, as expressed by human diploid cells (HDC) in vitro, is characterized by a progressive decline in proliferative activity. This phenomenon appears to be due both to a lengthening of the average cell cycle (Macieira-Coelho et al., '66; Absher et al., '74) and to a decrease in the rapidly proliferating fraction of the population (Cristofalo and Sharf, '73; Merz and Ross, '69). Previous studies (Cristofalo, '70, '73, '75) have shown that hydrocortisone (HC) at a concentration of 14 p M prolonged culture lifespan, i.e., population doubling level (PDL) achieved, by approximately 3040%. Macieira-Coelho ('66) and Smith et al. ('73) have also reported that various glucocorticoids could enhance the growth of normal HDC derived from embryonic lung. At first it was difficult to reconcile these findings with other reports in the literature which indicated, in general, that proliferative activity was inhibited by glucocorticoids (Ruhman J. CELL. PHYSIOL., 90: 415-422.

and Berliner, '65; Epifanova, '72). However, in a study of the effect of HC on other cell lines, we found that we could duplicate the inhibitory effects reported by others while enhancing the proliferation of human fetal lung cell lines under identical conditions (Cristofalo, '73). Thus the HC effect on HDC lifespan represents the action of a chemically defined modulator of cell proliferation and therefore provides a useful probe for exploring the regulation of cell proliferation. In order to better understand this effect, the cytokinetics of aging HDC cultures carried in parallel with or without the addition of the steroid were examined. The results of these studies comprise the basis for this report. MATERIALS AND METHODS

All studies were done with HDC strain WI-38 which was obtained either from Received Mar. 24, '76. Accepted July 1, '76.

415

416

GARY L . GROVE A N D V I N C E N T J. CRISTOFALO

frozen stock maintained at the Wistar In- curves. The cell cycle parameters were stitute or from Dr. Leonard Hayflick of derived using the Mendelsohn and Stanford University. Early passage (PDL Takahashi ('71) method of manual 18-21) starter cultures were divided into analyses, which provides estimates of the two substrains, both of which were cul- mean duration of the cell cycle phases, as tured as described by Cristofalo and well as the variation in cell cycle time. Sharf ('73) except that 14 pM HC was RESULTS added at each subcultivation to one of the For convenience, results from a single substrains. Proliferative activity was monitored longitudinal study are presented in this throughout the lifespan, using the au- paper; however, these are representative toradiographic method of Cristofalo and of what has been obtained in replicate Sharf ('73).Cultures in log phase growth studies of the effects of aging and HC on were continuously exposed to 3H-TdR at the cell cycle of human diploid cell line 0.1 pCi/ml (spec. act. 2.0 Ci/mM) for 30 WI-38. Figure 1 shows the percentage of cells hours. Prior to autoradiographic processing, the cells were stained by a modified able to incorporate 3H-TdR during a 30Feulgen procedure (Grove and Mitchell, hour labeling period as a function of '74). To estimate the size of the actively calendar age of the culture. In both subproliferating pool, the proportion of strains, an age-related decline in the fraclabeled cells was scored by counting tion of unlabeled cells could be ob1,000 cells. A total of 100 unlabeled cells served. However, the changes were dewere then selected at random and the rel- layed in the HC-supplemented cultures, ative amounts of Feulgen DNA per nu- which continued to proliferate after the cleus were determined with a Vickers control cultures could no longer be pasM85 integrating microspectrophotometer saged. at 560 nm. Results of the microspectrophotometThe duration of the interphase periods ric analyses of the cells that remained unof the cell cycle was determined by the labeled after 30 hours of continuous expercent labeled mitosis (PLM) method of posure to 3H-TdR are shown in table 1. Quastler and Sherman ('59) as modified All unlabeled cells in the early passage by Macieira-Coelho et al. ('66).Cultures cultures had a 2C amount of DNA, indiin log phase growth were exposed to cating that these cells were in a 3H-TdR at 1 pCi/ml (spec. act. 2.0 prereplicative-postmitotic stage. With Ci/mM) for 20 minutes. After washing the advancing age there was an increase in monolayers twice with complete medium the number of cells having 4C and 8C supplemented with 10 puglml nonlabeled amounts of DNA. Again, the two subTdR, fresh complete medium was added strains showed similar patterns of change to each culture. All media were pre- that were delayed in the presence of HC. warmed to 37°C and equilibrated with a Figure 2 shows the PLM curves for 5%CO, atmosphere prior to being added both control and HC-supplemented culto the cultures. In the case of HC-treated tures at various times after the initial adcells, the hormone was present at 14 p M dition of the steroid. At three weeks, the throughout the entire series of manipula- PLM curves of the two substrains are estions. Duplicate cultures were harvested sentially the same, with both displaying at 2-hour intervals over the next 30 hours. an initial peak having a plateau at the Autoradiographs were prepared and a 100% level and a prominent second minimum of 100 mitoses throughout the wave. At nine weeks, the curves are coverslips were scored for each sample. much the same as at three weeks, except The mean PLM were plotted against that the height achieved by the second time-after-pulse to generate the PLM wave is somewhat lower than that previ-

417

C E L L CYCLE O F HUMAN D I P L O I D C E L L S

CONTROL

HYDROCOR TISONE

2

6

4

8

10

12

14

16

18

20

24

22

WEEKS AFTER HYDROCORTISONE ADDITION Fig. 1 The effect of HC on the fraction of labeled cells in log phase cultures after 30 hours continuous exposure to 3H-TdR (0.1 pCiiml, 2 Ci/mM) as a function of weeks after the initial addition of the steroid. TABLE I

The effect of hydrocortisone on the DNA content of unlabeled Wl-38 cells Control cells Time after HC addition

3 weeks 9 weeks 15 weeks 21 weeks

HC-treated cells

Distribution3 Percent P.D.L.' unlabeledz 2C 4C 8C

30 50 60

-

4 17 44

-

100 91 73

-

0 7 22

-

0

2 5

-

P.D.L.'

34 62 73 82

Distribution3 Percent .~~.~~~~ unlabeledZ 2C 4C 8C

3 12 22 51

100 97 86 68

0 2 10 25

0 1 4

7

' Based on direct cell counts of cell yields at each week. Based on measurements of 1,000 cells. Based on Feulgen DNA measurement of 100 unlabeled nuclei.

ously achieved by the 3-week cultures, lack of a plateau at 100%suggest the exissuggesting that changes in G, have oc- tence of a delay in G,. Since the breadth curred. By 15 weeks the curves are of the initial peak is the same, no appreclearly different. The HC-treated cul- ciable change in the median duration of S tures still exhibit a fairly well-defined has occurred. Also, the second peak is sePLM curve. In contrast, the late passage verely damped, an indication that further control PLM curve is less well-defined, changes in the duration of G , have ocindicating that the intercellular variation curred. The control cultures could not be in cell cycle times within the population analyzed after the eighteenth week of has increased. The difference in the ini- the experiment; however, the magnitude tial peak with regard both to the steep- of lifespan extension was such that ness of the ascending slope and to the at 21 weeks measurements of HC-

418

GARY L. GROVE AND V I N C EN T J. CRI S TO F A LO

I

3 WEEKS

0

HYDROCORTISONE

2 rn

0

I-

5

n W -I

w m Q

-1

I-

2 W

0

a W n

Fig. 2 PLM of control and HC-treated substrains at various periods after the initial addition of 14 p M HC. PLM curves were generated by plotting the percentage of labeled mitoses in log phase cultures as a function of time after a 20-minute pulse exposure to 3H-TdR (1 p,Ci/ml, 2 Ci/mM).

supplemented cultures could still be obtained. The PLM curve obtained was poorly defined, indicating that changes in GI and G, also had occurred in these cultures. The data on parameters of the cell

cycle measured from the curves shown in figure 2 are presented in table 2. In control cultures early passage cells have an average cell cycle time of 19 hours, while late passage cells have an average cell cycle time of almost 31 hours. This

419

C E L L CYCLE O F HUMAN D I P L O I D C E L L S TABLE 2'

The effect of hydrocortisone on cell cycle parameters during aging in WI-38 cells Weeks after HC addition

3 9 15 21

Control tl

tz

ts

4.1 10.4 4.2 10.6 14.8 10.5

4.5 4.6 5.4

HC

cv

t

19.0 29.6 19.4 32.0 30.7 37.0

t,

t,

t2

t,

cv

35 4.8 7.5 18.0

10.5 10.3 10.7 10.9

4.4 4.6 4.6 5.8

18.4 19.7 22.8 34.7

22.7 29.6 28.9 48.9

' P h a s e durations in hours as derived from the PLM curves by the method of Mendelsohn and Takahashi ('71). TI, T,, T, and T, are respectively the mean durations of G, + 0.3mitosis, S, Gz + 0.7 mitosis and the entire cell cycle. CV is an approximation of the relative standard deviation of T,.

TABLE 3

Summary of cell cycle phase durations of normal embryonic human diploid fibroblastic cells Cell type

Early control 18th-35th p WI-26 ? HDC 18th p WI-38 18th-35th p WI-26 20th-24th p WI-38 Phase I1 LHC 18th p WI-38

t,

t

t

t

References

4.1 4.5-8 4.5-8 6.5 2.5 4.5 2.4-4.6 3.0

10.4 7-7.5 7.5 6 11.5 12.0 10.15 10.5

4.5 3-6 4 4.5 4.5 s3.5 5.4 3.9

19.0 18 18 17 18.5 19.5 18-20 17.4

This study Moorhead and Defendi, '63 Defendi and Manson, '63 Macieira-Coelho et al., '66 Cleaver, '67 Klevecz and Kapp, '73 Kukharenko et al., '74 Yanishevsky et al., '74

DISCUSSION change in cell cycle time is due almost Continuous 3H-TdR labeling, microentirely to changes in TI (duration of GI + 0.3 M). No appreciable change in the spectrophotometry and PLM analyses all median duration of S was observed. In were used to evaluate the effect of HC on late passage cultures a slight delay in T, the cytokinetics of aging HDC in vitro. (G, + 0.7 M) was observed. The coeffi- Identification of the actively proliferating cient of variation (CV) indicates that in- pool was done operationally by 30 hours tercellular variation in cell cycle times is of continuous exposure to 3H-TdR. This a characteristic of normal diploid human approach may not give an accurate meafibroblastic cultures, that becomes even sure of the growth fraction (i.e., number more pronounced in late passage popula- of cycling cells), particularly in older tions. In the HC-supplemented sub- populations, because slowly cycling cells strain, the average cell cycle time in- that have a longer cycle duration may not creased from 18.4 hours to 35 hours, due be labeled during this time. However, in almost entirely to changes in TI. The du- this manner we have found that with adration of S was the same as in control cul- vancing age there is a decrease in the tures, while a slight delay in T, occurred number of actively proliferating cells in in late passage cells near the end of their both control and HC-supplemented sublifespan. Thus, although the same pattern strains. The delay observed in the presof change occurs in both substrains, the ence of the hormone confirms the earlier changes are somewhat delayed in the observations of Cristofalo ('73) and supHC-treated substrains. ports the concept that the hormone-

420

GARY L. G R O V E A N D V I N C E N T J. CRISTOFALO

mediated enhancement of lifespan seems to be due, in part at least, to maintaining a higher percentage of the cells in the actively proliferating pool. The majority of unlabeled cells were found to be in the 2C Feulgen DNA mode, indicating that these cells were arrested in or slowly traversing G,. In late passage cultures an increase in the number of unlabeled 4C cells was also discerned. Whether this is due to tetraploid cells in G, or diploid cells in Gz cannot be resolved in this manner, but a concomitant increase in the degree of polyploidy (8C) in late passage cultures does suggest that a part of this fraction may be tetraploid cells in G,. These findings are in agreement with those previously reported by Grove and Mitchell ('74) and by Yanishevsky et al. ('74). Additional studies by Yanishevsky and Carran0 ('75), which used premature chromosome condensation, has shown that typical G, configurations were observed only among the labeled cells, while some GI tetraploid configurations were unlabeled. However, as pointed out by these workers, a possibility exists that unlabeled diploid cells in G, may not be detectable by this technique. Indeed, the previous cytokinetic studies of Macieira-Coelho et al. ('66), as well as data obtained in this study, indicate that large passage WI-38 cultures contain cells that have prolonged G, periods. Cytokinetic analysis of WI-38 and related human embryonic cell strains by the PLM method has been previously performed by a number of investigators. Table 3 summarizes the results of these earlier studies and shows that the reported data appears contradictory. However, this apparent lack of agreement is due mainly to the choice of method of analysis rather than any inconsistencies in the actual data. Indeed several of the earlier PLM curves have been reevaluated using more recent methods of analysis. Such is the case for Moorhead and Defendi ('63) by Cleaver ('67) and Macieira-Coelho et al. ('66) b y Yani-

shevsky e t al. ('74). We have also reevaluated the previously published curves using the Mendelsohn-Takahashi ('71) method of manual analysis and have found all to be in general agreement with the values obtained for early passage cultures obtained in this study. It is important to note that the Klevecz and Kapp ('74) values have been confirmed by following the progression of a synchronized population throughout the cell cycle. This means that the method of analysis used in the present study give a good approximation of the actual duration of the various cell cycle periods. Macieira-Coelho et al. ('66) previously reported that the cell cycle becomes prolonged and more variable with advancing culture age. The present study confirms and extends their previous study by showing that similar changes also occur in cultures whose lifespan has been extended by HC. Since the duration of mitosis does not change appreciably with age (Macieira-Coelho et al., '66), the changes in T, and T, reflect primarily changes in the duration of GI and G,, respectively. Thus the cell cycle duration is prolonged in late passage cultures primarily due to changes in G,. A slight retardation in G, also was observed in late passage cultures. These changes are similar to the age-related changes reported by Macieira-Coelho et al. ('66). Our results also show that the duration of DNA synthesis does not change with advancing age. This finding is in contrast with the report of Macieira-Coelho et al. ('66), but are in agreement with the more recent reevaluations of the earlier date (Yanishevsky et al., '74; Macieira-Coelho et al., '75). A number of models describing the kinetics of cell proliferation in various aging populations have been described. Among such models are those proposed by Good ('72), Gelfant and Smith ('72), Gelfant and Grove ('74) and Yanishevsky et al. ('74). These models share much in common and postulate the existence of distinct populations of "cycling" and

C E L L CYCLE OF H U M A N DIPLOID CELLS

"noncycling" cells. In these models, changes in proliferative activity are simply explained in terms of cycling and noncycling transitions, while the cell cycle is assumed to be of constant duration. However, the PLM curves obtained in this study illustrate the extent of the changes in the cell cycle which occur with advancing age. The present results may be related more to a cell cycle model devised by Smith and Martin ('73, '74), who proposed that the cell cycle is composed of an A-state and a B-phase. After mitosis each cell enters the A-state where it can remain for an indeterminate period. The B-phase consists of the conventional s, G, and M phases, as well as of some part of GI and is of determinate duration. Entrance of cells from the A-state to the B-phase is dependent on the transition probability (P), and it is the modification of this value that provides the major means of controlling proliferative activity. Burns and Tannock ('72)and De Maertalaer and Galand ('75) have proposed similar models to describe the behavior of proliferative populations in which cells can halt temporarily in some part of GI. It should be emphasized that the duration of GI includes the time spent in the A-state. However, to encompass the G, delay and the polyploidy observed in late passage cultures, these models require some modification. Nevertheless, HDC cultures should be considered as having a spectrum of cell types, including rapidly proliferating cells where P is high approaching unity, slowly proliferating cells where P is somewhat lower, and nonproliferating cells where P approaches zero. Our viewpoint is consistent with the notion that cellular senescence is due to a transition from a rapidly proliferating state through a series of more slowly cycling ones, and finally to a state in which the cells are arrested or cycling so slowly as to be unable to repopulate the vessel (Cristofalo, '75.; Macieira-Coelho et al., ' 7 5 ) . One simple interpretation of the HC extension of culture lifespan is that the steroid

42 1

somehow delays these transitions. On the basis of the information obtained in this study, it seems likely that the site of action of this hormone resides in the preDNA synthetic state. Indeed, preliminary studies (Grove and Cristofalo, '76) indicate that HC need only be present during the first six hours of the pre-DNA synthetic period to enhance the number of fibroblasts stimulated to proliferate by media renewal in confluent cultures. It is interesting to note that in HeLa cultures HC decreases the rate of cell proliferation by considerably extending the generation time and that this is entirely due to changes in the duration of the preDNA synthetic phase (Kollmorgen and Griffin, '69). Our knowledge concerning the hormonal regulation of cell proliferation is still incomplete. We feel that an understanding of the hydrocortisone-diploid cell interaction is probably important to our understanding of the mechanism of glucocorticoid hormones in general as well as to our understanding of proliferation and senescence in vitro. ACKNOWLEDGMENTS

We wish to thank Dr. A. MacieriaCoelho for his critical review of the manuscript. The excellent technical assistance of Joanne Cochran and Dori Green is gratefully appreciated. This investigation was supported by USPHS Research Grant AGO0378 from the National Institute of Aging. LITERATURE CITED Absher, P. M., R. G. Absher and W. D. Barnes 1974 Geneologies of clones of diploid fibroblasts: cinemicrophotographic observations of' cell division patterns in relation to population age. Exptl. Cell Res., 88: 95-104. Burns, F. J., and I. F. Tannock 1970 On the existence of a Go-phase in the cell cycle. Cell Tissue Kinet., 3: 321-334. Cleaver, J. E. 1967 Thymidine metabolism and cell kinetics. North Holland Publishing Co., Amsterdam, 126 pp. Cristofalo, V. J. 1970 Metabolic aspects of aging in diploid human cells. In: Aging in Cell Tissue Culture. E. Holeckova and V. Cristofalo, eds. Plenum Press, New York, pp. 83-119.

422

GARY L. G R O V E AND V I N C E N T J . CRISTOFALO

-

1973 Cellular senescence: factors modulating cell proliferation in vitro. Mol. and Cell. Mechanisms of Aging, INSERM, 27: 65-92. 1975 The effect of hydrocortisone on DNA synthesis and cell division during aging in uitro. In: Cell Impairment in Aging and Development. V. J. Cristofalo and E. HolE-ckova, eds. Plenum Press, New York, pp. 7-22. Cristofalo, V. J., and B. B. Sharf 1973 Cellular senescence and DNA synthesis: thymidine incorporation as a measure of population age in human diploid cells. Exptl. Cell Res., 76: 419-427. Defendi, V., and L. A. Manson 1963 Analysis of the life-cvcle in mammalian cells. Nature, 198: 356-361. De Maertelaer. V.. and P. Galand 1975 Some properties of a “Go” model of the cell cycle. I. Investigations of the possible existence of natural constraints on the theoretical model in steady-state conditions. Cell Tissue Kinet., 8: 11-22. Epifanova, 0. I. 1972 Effects of hormones on the cell cycle. In: Cell Cycle and Cancer. R. Baserga, ed. Marcel Dekker, New York, pp. 145-190. Gelfant, S. and G. L. Grove 1974 Cycling noncycling cells as an explanation for the aging process. In: Theoretical Aspects of Aging. M. Rockstein, ed. Academic Press, New York, pp. 105-117. Gelfant, S., and J. G. Smith, Jr. 1972 Aging: noncycling cells an explanation. Science, 178: 357-361. Good, P. 1972 Homeostatic mechanisms and the loss of mitotic potential. J. Theor. Biol., 34: 99-102. Grove, G. L., and V. J. Cristofalo 1976 Effects of hydrocortisone on WI-38 fibroblasts stimulated to proliferate. In Vitro, 12: 299. Grove, G. L., and R. B. Mitchell 1974 DNA microdensitometry as a measure of cyclingnoncycling activity in aged human diploid cells in culture. Mech. Age. Dev., 3: 235-240. Klevecz, R. R., and L. N. Kapp 1973 Intermittent DNA synthesis and periodic expression of enzyme activity in the cell cycle of WI-38. J. Cell Biol., 58: 564-573. Kollmorgen, G. M., and M. J. Griffin 1969 The effect of hydrocortisone on HeLa cell growth. Cell Tissue Kinet., 2: 111-122.

-

,

I

*

Kukharenko, U. I., A. M. Kuliev, K. N. Grinberg and V. V. Terokikh 1974 Cell cycles in human diploid and aneuploid strains. Humangenetik, 24: 285-296. Macieira-Coelho, A., E. Loria and L. Berumen 1975 Relationship between cell kinetic changes and metabolic events during cell senescence in uitro. In: Cell Impairment in Aging and Development. V. J. Cristofalo and E. Holeckova, eds. Plenum Press, New York, pp. 51-65. Macieira-Coelho, A., J. Ponten and L. Philipson 1966 The division cycle and RNA synthesis in diploid human cells at different passage levels in uitro. Exptl. Cell Res., 42: 673-684. Mendelsohn, M. L., and M. Takahashi 1971 A critical evaluation of the fraction of labeled mitoses method as applied to the analysis of tumor and other cell cycles. In: The Cell Cycle and Cancer. R. Baserga, ed. Marcel Dekker, Inc., New York, pp. 55-95. Merz, G . ,and J. Floss 1973 Clone size variation in the human diploid cell strain, WI-38. J . Cell Physiol., 82: 75-80. Moorhead, P. S., and V. Defendi 1963 Asynchrony of DNA synthesis in chromosomes of human diploid cells. J . Cell Biol., 16: 202-209. Quastler, H., and F. G . Sherman 1959 Cell population kinetics in the intestinal epithelium of the mouse. Exptl. Cell Res., 17: 420-438. Ruhman, A. G., and D. L. Berliner 1965 Effect of steroids on growth of mouse fibroblasts in vitro. Endocrinol., 76: 916-927. Smith, B. T., J. S. Torday and C. S. P. Giroud 1973 The growth promoting effect of coritsol on human fetal lung cells. Steroids, 22: 515-523. Smith, J. A., and L. Martin 1973 Do cells cycle? Proc. Nat. Acad. Sci. (U.S.A.), 70: 1263-1267. - 1974 Regulation of cell proliferation. In: Cell Cycle Controls. G . M. Padilla, I. L. Cameron and A. Zimmerman, eds. Academic Press, New York, pp. 43-60. Yanishevsky, R., and A. V. Carrano 1975 Prematurely condensed chromosomes of dividing and non-dividing cells in aging human cell cultures. Exptl. Cell Res., 90: 169-174. Yanishevsky, R., M. L. Mendelsohn, B. H. Mayall and V. J. Cristofalo 1974 Proliferative capacity and DNA content of aging human diploid cells in culture: a cytophotometric and autoradiographic analysis. J. Cell Physiol., 84: 165-170.

Characterization of the cell cycle of cultured human diploid cells: effects of aging and hydrocortisone.

Characterization of the Cell Cycle of Cultured Human Diploid Cells: Effects of Aging and Hydrocortisone GARY L. GROVE AND VINCENT J. CRISTOFALO The Wi...
599KB Sizes 0 Downloads 0 Views