Mechanisms of Ageing and Development, 4 (1975) 289-299 © Elsevier Sequoia S.A., Lausanne - Printed in The Netherlands
289
T H E L A C K OF A L T E R E D E N Z Y M E M O L E C U L E S I N " S E N E S C E N T " M O U S E E M B R Y O FIBROBLASTS I N C U L T U R E
M. DANOT and H. GERSHON* Department of Immunology, Aba Khoushy Medical School, Bat Galim, Technion, Haifa (Israel) D. GERSHON Department of Biology, Technion, Haifa (Israel) (Received June 25, 1975)
SUMMARY In order to compare the biochemical events occurring in senescent animals in vivo with those o f fibroblasts demonstrating a limited life span in tissue culture, experiments were performed to determine whether altered enzyme molecules are produced in late passage mouse embryo fibroblasts in tissue culture. Such altered enzymes have been previously described by us in both senescent mice and nematodes. Fibroblast cultures, initiated from BALB/c embryos underwent approximately 7 population doublings before reaching stationary phase. A comparison was made of the enzyme activity and the quantity of antigen of the enzyme aldolase present in these cultures. The amount of aldolase per cell as well as the amount of enzyme per unit of antigen remained constant throughout the in vitro lifespan of these cells. Further evidence testifying to a lack of enzyme alteration in vitro was obtained in experiments testing temperature sensitivity. The aldolase from all passages was equally sensitive to heating at 60°C. INTRODUCTION The experiments of Hayflick and Moorhead 1 and Hayflick 2 and subsequently of others, have established a new doctrine relating to cells in tissue culture. These findings imply, in negation to the original findings of Carrel 3, that normal, diploid fibroblasts can undergo but a limited number of cell divisions in vitro, after which, they lose their viability and the cultures degenerate. It has also been reported that the number of population doublings which such fibroblasts can undergo is inversely proportional to the age of the original donor2, 4.*. The number of population doublings is also claimed to be related to the maximal expected life span of the fibroblast donor species, as demonstrated in a comparison of human, chicken and mouse fibroblastsL * To whom all correspondence should be addressed. ** Editorial note: The scatter between results from individual cultures as a function of age is appreciable, however, and the degree of decline in doublings in post-natally derived cultures is not impressive.
290 These findings of mortality of cells in vitro have led to the possibility that diploid cells in tissue culture may be a suitable model for studies of aging, and that the conditions and phenomena responsible for the termination of the in vitro life span of fibroblasts may in fact be the same as those responsible for the senescense and death of the intact organism. In studies of the biochemical basis of senescence and death in vivo, our laboratory has investigated the possibility that altered protein molecules may appear during the aging process 6. In the enzymes isocitrate lyase and aldolases from then ematode Turbatrix aceti 7,8 and in both the muscle and liver aldolase of C57B! miceg, ~0, a significant drop in enzyme specific activity with age was demonstrated, while enz2cme antigen remained constant. These and other experiments have led us to conclude that catalytically inactive, antigenically cross reactive (CRM) enzyme molecules accumulate as a function of age. The fact that age-dependent CRM appears in two such phylogenetically disparate species as the nematode and the mouse leads one to speculate that the biochemical processes leading to CRM production may, indeed, be closely related to the causes of senescence. In the light of these findings, we thought it important to determine whether faulty enzymes are produced as a function of time in culture of fibroblasts which exhibit a limited life span in vitro. In order to obtain information which would parallel our in vivo experimentation in mice, studies were performed on the enzyme aldolase which is present in cultures of fibroblasts obtained from mouse embryos. MATERIALS AND METHODS Tissue culture
Fibroblasts were obtained from 12-14 day old BALB/c embryos and cultured in Dulbecco Modified Eagles Minimal Essential Medium (H-16, Gibco) plus 10% calf serum. The initial concentration of cells in the primary cultures was 5 × 106 embryonic cells per 60 mm plastic tissue culture plate (Falcon). Media in all plates were changed every 2-3 days. Cultures were split I :2, by trypsinization upon reaching confluency (every 4 days in the early passages and 5-7 days in late passages). These cultures reached stationary phase after six to seven passages beyond which confluency was not attained even after two weeks in culture or longer. Under the same conditions chick embryo fibroblasts undergo 27 ± 3 doublings (D. Gershon, unpublished results), a figure similar to that reported by Lima and Maciera-Coelho 11 and Hay, Menzies, Morgan and Strehler 12. Cell homogenization
Cells from individual tissue culture plates were harvested with a rubber policeman and transferred into 1 ml of 50mM Tris HCI, 10mM /3-mercaptoethanol, 1 mM E D T A buffer (50mM Tris buffer), homogenized at 4°C with a Dounce homogenizer and further disrupted by sonication for 2 minutes to release aldolase. The homogenate was centrifuged at 30 000 × g for 30 minutes at 4°C. The resulting supernatant was maintained at 4°C and used in subsequent assays of enzyme activity.
291 Aldolase assay Aldolase activity was assayed as described by Gracy, Lacko, Brox, Adelman and Horeoker 13 with the exception that 15/~g of glycerol phosphate dehydrogenasetriose phosphate isomerase mixture (Boehringer, Mannheim Corp.) was substituted for 10#g of each of the individual enzymes. One unit of aldolase activity is defined as the amount of enzyme catalyzing the cleavage of 1 #mole of fructose 1.6 diphosphate per minute at 25°C. Protein determinations Protein determinations were performed according to the method of Lowry, Rosebrough, Farr and Randall 14. Verification of cell normality To ascertain that no malignant transformation had occurred during in vitro passaging of the mouse fibroblasts, these cells were reinjected into isologous 2 month old recipients. Each recipient was injected intramuscularly with 4 × 106 cells from either the first passage or seventh passage (stationary phase). Each recipient mouse was examined for palpable tumors every fortnight for three months with no evidence of tumor growth in any of the mice injected. The diploid chromosome number was verified in cells of each passage as described by Merchant, Kahn and MurphylL A single exception to the recommended protocol was that cultures were exposed to colchicine for 16 rather than 8 hours. All preparations provided evidence that at each in vitro passage, the fibroblasts bore a normal diploid number of chromosomes per cell. Antisera to aldolase Monospecific antisera to purified mouse liver and mouse muscle aldolase from 3-4 month old mice were prepared as previously described 9,1°. The amount of aldolase antigen in each tissue culture homogenate was assayed by mixing a fixed amount o f homogenate (200/zl) with varying volumes of antiserum plus 50mM Tris buffer to make 250 #1 incubation mixture and incubated at 4°C overnight and then centrifuged at 3000 × g for 30 minutes at 4°C. The precipitated antigen-antibody complex was discarded and 100/zl of the supernatant was assayed for residual enzyme activity. Controls of both buffer and normal rabbit serum were always included. Serum alone showed low aldolase activity and this was subtracted from the recorded aldolase activity. RESULTS The growth parameters of the BALB/c mouse embryo fibroblasts in tissue culture can be seen in Fig. 1. By the sixth passage these cells have greatly slowed down in their ability to replicate and by the seventh transfer, they have virtually stopped growing. Preliminary experiments determined that the aldolase in the fibroblast cultures
292 Passages
I
-e--A--
]I 1I
5 -
--D--
E "Vr
2
{0
$ O
Z
I l
I 2
3
I 4
I 5
Days in culture
Fig. 1. Growth of mouse embryo fibroblasts as a function of passage in vitro. Cells were harvested by trypsinization and viable cell counts were performed with the aid of 0.05 ~ trypan blue in phosphate buffered saline.
is aldolase A (muscle type). Monospecific rabbit anti-mouse aldolase A both inactivated and precipitated this enzyme whereas monospecific rabbit anti-mouse aldolase B (river type) had virtually no effect on the aldolase activity of the culture homogenates. Determinations of enzyme activity per day in culture per passage showed that the specific activity of aldolase per mg of protein does not vary significantly from passage to passage, as seen in Fig. 2. Because the protein content of these
293 Passages
-a-
!
--&--
11'
-o-
11"
--II--
~"
--13-
"~
Z2
64
5.6
4.8 A
c
n o) "~E3.2 _ c !:3 "~ 2.4
~
1.6
I
I !
I 2
I 3
I 4
I 5
D a y s in c u l t u r e
Fig. 2. The specific activity of Aldolase A in the homogenates of individual cultures as function of time in culture. cultures decreased in proportion to cell number, this constant specific activity reflects a constant amount of aldolase activity per cell in each passage. A constant specific activity does not necessarily represent a constant level o f enzyme. There may well be inactive molecules (or molecules with reduced specific activity) present during different phases o f growth in vitro. In order to compare the amount o f aldolase antigen in each passage, cultures were harvested for the titration of enzyme by antibody three to four days after the initiation of each passage. Fig. 3 is a typical precipitation curve. In order to take advantage of the linear section of the precipitation curve, the quantity of enzyme precipitated by 1 F1 of antiserum
294
eQ
1.c e,-
~..~point of 50z precipitation
. i
Qe i
Io
0.2
I
I 2
I 4
I 6
I 8
I 10
Quantity of Antiserum (,ul)
I 12
Fig. 3. The precipitation of Aldolase A by antibody: A typical titration. The following calculation of units of enzyme precipitated by 1 #1 of antibody at the point of 50~ precipitation was made: Units of enzyme precipitated by 1/~l antiserum = units of enzyme precipitated at 50 ~ precipitation/ microliters of antisera required to precipitate 50 ~ activity was calculated from the point of 50 ~ precipitation. This calculation of the amount of enzyme precipitated by 1 #1 of antiserum with antigen in excess is valid over a wide range of initial enzyme activities, as shown in Fig. 4, and is, therefore, valid for comparing enzyme preparations which may contain sequestered enzyme present in the form of CRM. Comparison of the amount of enzyme activity precipitated by 1/A of antiserum on day 3 or 4 of each passage shows that the amount of enzyme antigen remains constant throughout the in vitro lifespan of mouse embryo fibroblasts (Fig. 5). Even though the specific enzyme activity per number of cells remains at a constant level, and there is no detectable evidence of an altered amount of enzyme antigen per successive passage in vitro certain minor alterations in enzyme structure might nevertheless occur which, under normal physiological conditions, might not express themselves in either altered enzyme activity or conformation. Such alterations might, however, express themselves under denaturing conditions e.g. heat inactivation. Homogenates of 3 or 4 day old cultures from each passage were, therefore, subjected to 60°C in a constant temperature water bath and the residual enzyme activity was assayed after various times of exposure. No differences in temperature sensitivity were observed over the entire #7 vitro life span of these fibroblasts. Figure 6 depicts the heat inactivation of enzyme from cells of passages I (secondaries) and VI. Enzyme from all intermediary passages showed similar heat sensitivity. Such a non linear (lst order) pattern of inactivation may reflect either a molecular heterogeneity
295
E "5
5"" Q4q
~
~..t•
,_,.,..gp•
"~Q24
"~
I
O
I
03
Initial
1,6
Aldolase
I
2.4
Activity (milliunil"s)
Fig. 4. The amount of enzyme activity precipitated by 1/d of antibody at 50% precipitation as a function of initial enzyme activity. Three secondary cultures were harvested and pooled before disruption. The resulting homogenate was assayed for aldolase activity and serial dilutions were made in 50 mM Tris Buffer to give aliquots with different initial aldolase activities. Each aliquot was titrated with antibody and the amount of enzyme precipitated by 1 /d of antiserum at 50% precipitation was performed as described in Fig. 3.
o f the aldolase which does not vary t h r o u g h o u t the in vitro lifespan, or selective heat sensitivity o f certain sites on the aldolase A molecule, while others are more resistant. DISCUSSION BALB/c mouse e m b r y o fibroblast cultures undergo a limited n u m b e r o f population doublings in vitro and then cease to replicate. In this respect mouse embryonic fibroblasts resemble fibroblasts o f other animal speciesZ,4, al. The p r o p o r t i o n o f cells capable o f synthesizing D N A in the mouse cultures decreases from 95 % in the first passage to 4 0 % in the last passage before the onset o f mitotic quiescence16; these results are similar to those reported for h u m a n fibroblasts whose life-span in vitro is limited17, a8 to about 50-70 doublings. We are aware that mouse embryo cells have been reported to show a high degree o f " s p o n t a n e o u s " neoplastic transformation when cultivated in vitro. However, when one scrutinizes carefully the data presented in papers reporting this p h e n o m e n o n it becomes obvious that this transformation and the subsequent f o r m a t i o n o f cell lines takes a few months to appear 19-z2. It is usually detected after an initial growth phase which is followed by a long quiescent period. The renewed growth after such a prolonged quiescent period is most likely due to a selection o f a small n u m b e r o f
296 E 2 E
.D ~(~0.64
0
~o.,~ t~
E N>'0.32 ¢...
u) 0.16 .~_ (.I
I
-n-
1
I
~I~ -I1Passages in culture
I
-7
I
I
Fig. 5. The amount of enzyme activity precipitated by 1 td of antiserum at 50% precipitation as a function of passage in culture. Cultures of passages I-III were assayed on day 3. Cultures of passages IV-VI were assayed on day 4.
transformed cells. It is unclear whether these cells appear in minute numbers during the growth phase or in the subsequent quiescent phase. Certainly growth conditions such as serum type 19-22 and cell density za exert a profound effect on the rate of transformation. There are several causes which one might suggest to explain the occurrence of this phenomenon of a limited life span of fibroblasts in vitro. One such suggested cause might be the production of altered enzyme molecules in the course of in vitro replications. Such enzyme molecules might result in an "error catostrophe" such as that suggested by Orgel z4,25 and, therefore, result in the death of the cultures. Alternatively, altered protein molecules may have a less drastic effect which could incapacitate the cell sufficiently to slow down its growth rate and eventually cause it to cease replication. Our findings of similar heat stability and a constant ratio of enzyme activity to enzyme antigen do not support the presence of altered enzyme molecules in mouse embryo fibroblast cultures which, nevertheless, posses a constant, limited in vitro lifespan. These findings are in contrast to parallel in vivo systems which have been reported from our laboratory 7-10 for in vivo experiments carried out both on nematodes and on mice which clearly reveal the accumulation of inactive enzyme molecules as a function of increasing age. Our results differ from those reported for human embryonic fibroblasts of the MRC-5 strain in which an accumulation of large amounts of altered protein leading to an "error catastrophe" has been reported. Lewis and Tarrant z~ have reported the
297
+I 8O
,I-
.5
I0
15
20
25
30
Time (minutes)
Fig. 6. The thermobility of Aldolase A from fibroblasts of in vitro passages I and VI. Culture homogenates were exposed to 60 °C in a constant temperature water bath. Aliquots were removed for assay of aldolase activity at the time intervals denoted above.
presence o f C R M to the enzyme, lactic dehydrogenase. This enzyme has 5 isozymes and even t h o u g h it has been reported that the relative amounts o f each isozyme remain constant t h r o u g h o u t the cultured lifespan o f these cells, interpretation o f such results is very difficult, especially when one is dealing with an apparently exceedingly abrupt decrease in detectable enzyme activity. Holliday and Tarrant 27 have studied the heat sensitivity o f the enzyme glucose-6-phosphate dehydrogenase (G-6-PD) and report an increase o f 20-25 ~ heat labile enzyme in late passage fibro-
298 blasts. In another publication, Holliday zs presents evidence that these same fibroblasts in culture undergo a selective process in which cells producing a new form o f G - 6 - P D become more and more frequent in culture. The selective advantage o f cells bearing a new isozymic form o f G - 6 - P D with, in all probability, a different sensitivity to heat denaturation might well account for the changes in heat stability observed in mass cultures. In contrast to the above findings, however, it should be noted that Martin and N o r w o o d (personal communication, T. N o r w o o d ) have not been able to detect any age-dependent C R M accumulation o f G-6-PD in WI-38 h u m a n fibroblasts in culture. Also, both Holland et al. 29 and Tomkins et al. 3° have employed viral probes for the detection of mis-synthesis o f proteins in h u m a n fibroblasts as a function o f age in culture. They showed that cells from early and terminal passages in culture were equally susceptible to herpesvirus type I, poliovirus type 129 and vesicular stomatitis virus30; the virus produced in cells o f all passages had the same specific activity and the senescent cells did not produce a larger p r o p o r t i o n o f defective virions. Holland et al. 29 also showed that viruses produced in " y o u n g " and "senescent" cultures have similar thermal stabilities and exhibit similar mutation rates. These results and the results reported in this communication indicate that whatever the cause o f the mortality o f fibroblasts in vitro it does not appear to be due to an error catasstrophe, nor, for that matter, to any general modification o f proteins. Relevant to this may be the recent (though as yet unconfirmed) findings o f Packer and Smith 31 which are consistent with the hypothesis that senescence o f fibroblasts in culture m a y be due primarily to oxidative damage. ACKNOWLEDGEMENTS This research has been supported by a grant from The Deutsche Forschungsgemeinschaft. We thank Mrs. Leah Rosenfelder for her skillful technical assistance.
REFERENCES 1 L. Hayflick and P. S. Moorhead, The serial cultivation of human diploid cell strains, Exp. Cell Res., 25 (1961) 585-621. 2 L. Hayflick, The limited in vitro lifetime of human diploid cell strains, Exp. Cell Res., 37 (1965) 614-636. 3 A. H. Ebeling, The permanent life of connective tssue outside of the organism, J. Exp. Med.. 17 (1913) 273-285. 4 G. M. Martin, C. A. Sprague and C. J. Epstein, Replicative life span of cultivated human ceils. Effects of donor's age, t!ssue and gcnotype, Lab. Invest., 23 (1970) 86-92. 5 L. Hayflick, The biology of human aging, Am. J. Med. Sci., 265 (1973) 432445. 6 H. Gershon, P. Zeelon and D. Gershon, Faulty proteins: Altered gene products in senescent cells and organisms. In A. Kohn and A. Shatkay (Eds.), ControlofGene Expression, Plenum, New York, 1974, pp. 255-264. 7 H. Gershon and D. Gershon, Detection of inactive enzyme molecules in aging organisms, Nature (London), 227 (1970) 1214-1217. 8 P. Zeelon, H. Gershon and D. Gershon, Inactive enzyme molecules in aging organisms: Nematode fructose-l,6-diphosphate aldolase, Biochemistry, 12 (1973) 1743-1750.
299 9 H. Gershon and D. Gershon, Altered enzyme molecules in senescent organisms: mouse muscle aldolase, Mech. Ageing Dev., 2 (1973) 3 3 4 1 . 10 H. Gershon and D. Gershon, Inactive enzyme molecules in aging mice: Liver aldolase, Proc. Natl. Acad. Sci. U.S.A., 70 (1973) 909-913. 11 L. Lima and A. Maciera-Coelho, Parameters of aging in chicken embryo fibroblasts cultivated in vitro, Exp. Cell Res., 70 (1972) 279-284. 12 R. J. Hay, R. A. Menzies, H. P. Morgan and B. L. Strehler, The division potential of cells in continous growth as compared to cell subcultivated after maintenance in stationary phase, Exp. Gerontol., 3 (1968) 3 5 4 4 . 13 R. W. Gracy, A. G. Lacko, L. W. Brox, R. C. Adelman and B. L. Horecker, Structural relations in aldolases purified from rat liver and muscle and Novikoff hepatoma, Arch. Biochem. Biophys., 136 (1970) 480-490. 14 O. H. Lowry, N. J. Rosebrough, A. L. Farr and R. T. Randall, Protein measurements with the Folin phenol reagent, J. Biol. Chem., 193 (1951) 265-275. 15 D. J. Merchant, R. H. Kalm and W. H. Murphy, Handbook o f Cell and Organ Culture, 2nd Ed. Burgess Publishing Co., Minneapolis, Minn., 1964. 16 M. Danot, Fetal mouse fibroblast cell cultures, a possible model for the study of aging? M.Sc. Thesis, Technion-Israel Institute of Technology, 1974. 17 A. Macieira-Coelho, Aging and Cell division, Front. Matrix, BioL, 1 (1973) 46-77. 18 V. J. Cristofalo and B. B. Sharf, Cellular senescence and D N A synthesis: thymidine incorporation as a measure of population age in h u m a n diploid cells, Exp. Cell. Res., 76 (1973) 419423. 19 K. K. Sanford, "Spontaneous" neoplastic transformation of cells in vitro: Some facts and theories, Nat. Cancer Inst. Monograph, 26 (1967) 387418. 20 K. K. Sanford, S. L. Handleman, J. W. Hartley, J. L. Jackson and R. R. Gnatt, Spontaneous neoplastic transformation in vitro: Influence of Endogenous murine leukemia virus and serum fractions, J. Natl. Cancer Inst., 49 (1972) 1177-1182. 21 V. J. Evans and W. F. Andersen, Effect of serum on spontaneous neoplastic transformations in vitro, J. Natl. Cancer lnst., 37 (1966) 247-249. 22 G. Carbone, R. Piazza and G. Parmiani, Effect of different sera on grwoth and "spontaneous" neoplastic transformation of mouse fibroblasts in vitro, J. Natl. Cancer Inst., 52 (1974) 387-393. 23 S. A. Aaronson, J. W. Hartely and G. J. Todano, Mouse leukemia virus : "Spontaneous" release by mouse embryo cells after long-term in vitro cultivation, Proc. Natl. Acad. Sci. U.S.A., 64 (1969) 87-94. 24 L. E. Orgel, The maintenance of the accuracy of protein synthesis and its relevance to aging, Proc. Natl. Acad. Sci. U.S.A., 49 (1963) 517-521. 25 L. E. Orgel, The maintenance of the accuracy of protein synthesis and its relevance to aging, Proc. Natl. Acad. Sci. U.S.A., 67 (1970) 1476. 26 C. M. Lewis, and G. M. Tarrant, Error theory and aging in human diploid fibroblasts, Nature (London), 239 (1972) 316-318. 27 R. Holliday and G. M. Tarrant, Altered enzymes in aging h u m a n fibroblasts, Nature (London), 238 (1972) 26-30. 28 R. Holliday, Humangenetik, 16 (1972) 83. 29 J. J. Holland, D. Kohne and M. V. Doyle, Analysis of virus replication in aging human fibroblast cultures, Nature (London), 245 (1973) 316-317. 30 G. A. Tomkins, E. J. Stanbridge and L. Hayflick, Viral probes of aging in the human diploid cell strain WI 38, Proc. Soc. Exp. Biol. Med., 146 (1974) 385-390. 31 L. Packer and J. R. Smith, Extension of the lifespan of cultured normal h u m a n diploid cells by vitamin E, Proc. Natl. Acad. Sci. U.S.A., 71 (1974) 4763-4767.
289
T H E L A C K OF A L T E R E D E N Z Y M E M O L E C U L E S I N " S E N E S C E N T " M O U S E E M B R Y O FIBROBLASTS I N C U L T U R E
M. DANOT and H. GERSHON* Department of Immunology, Aba Khoushy Medical School, Bat Galim, Technion, Haifa (Israel) D. GERSHON Department of Biology, Technion, Haifa (Israel) (Received June 25, 1975)
SUMMARY In order to compare the biochemical events occurring in senescent animals in vivo with those o f fibroblasts demonstrating a limited life span in tissue culture, experiments were performed to determine whether altered enzyme molecules are produced in late passage mouse embryo fibroblasts in tissue culture. Such altered enzymes have been previously described by us in both senescent mice and nematodes. Fibroblast cultures, initiated from BALB/c embryos underwent approximately 7 population doublings before reaching stationary phase. A comparison was made of the enzyme activity and the quantity of antigen of the enzyme aldolase present in these cultures. The amount of aldolase per cell as well as the amount of enzyme per unit of antigen remained constant throughout the in vitro lifespan of these cells. Further evidence testifying to a lack of enzyme alteration in vitro was obtained in experiments testing temperature sensitivity. The aldolase from all passages was equally sensitive to heating at 60°C. INTRODUCTION The experiments of Hayflick and Moorhead 1 and Hayflick 2 and subsequently of others, have established a new doctrine relating to cells in tissue culture. These findings imply, in negation to the original findings of Carrel 3, that normal, diploid fibroblasts can undergo but a limited number of cell divisions in vitro, after which, they lose their viability and the cultures degenerate. It has also been reported that the number of population doublings which such fibroblasts can undergo is inversely proportional to the age of the original donor2, 4.*. The number of population doublings is also claimed to be related to the maximal expected life span of the fibroblast donor species, as demonstrated in a comparison of human, chicken and mouse fibroblastsL * To whom all correspondence should be addressed. ** Editorial note: The scatter between results from individual cultures as a function of age is appreciable, however, and the degree of decline in doublings in post-natally derived cultures is not impressive.
290 These findings of mortality of cells in vitro have led to the possibility that diploid cells in tissue culture may be a suitable model for studies of aging, and that the conditions and phenomena responsible for the termination of the in vitro life span of fibroblasts may in fact be the same as those responsible for the senescense and death of the intact organism. In studies of the biochemical basis of senescence and death in vivo, our laboratory has investigated the possibility that altered protein molecules may appear during the aging process 6. In the enzymes isocitrate lyase and aldolases from then ematode Turbatrix aceti 7,8 and in both the muscle and liver aldolase of C57B! miceg, ~0, a significant drop in enzyme specific activity with age was demonstrated, while enz2cme antigen remained constant. These and other experiments have led us to conclude that catalytically inactive, antigenically cross reactive (CRM) enzyme molecules accumulate as a function of age. The fact that age-dependent CRM appears in two such phylogenetically disparate species as the nematode and the mouse leads one to speculate that the biochemical processes leading to CRM production may, indeed, be closely related to the causes of senescence. In the light of these findings, we thought it important to determine whether faulty enzymes are produced as a function of time in culture of fibroblasts which exhibit a limited life span in vitro. In order to obtain information which would parallel our in vivo experimentation in mice, studies were performed on the enzyme aldolase which is present in cultures of fibroblasts obtained from mouse embryos. MATERIALS AND METHODS Tissue culture
Fibroblasts were obtained from 12-14 day old BALB/c embryos and cultured in Dulbecco Modified Eagles Minimal Essential Medium (H-16, Gibco) plus 10% calf serum. The initial concentration of cells in the primary cultures was 5 × 106 embryonic cells per 60 mm plastic tissue culture plate (Falcon). Media in all plates were changed every 2-3 days. Cultures were split I :2, by trypsinization upon reaching confluency (every 4 days in the early passages and 5-7 days in late passages). These cultures reached stationary phase after six to seven passages beyond which confluency was not attained even after two weeks in culture or longer. Under the same conditions chick embryo fibroblasts undergo 27 ± 3 doublings (D. Gershon, unpublished results), a figure similar to that reported by Lima and Maciera-Coelho 11 and Hay, Menzies, Morgan and Strehler 12. Cell homogenization
Cells from individual tissue culture plates were harvested with a rubber policeman and transferred into 1 ml of 50mM Tris HCI, 10mM /3-mercaptoethanol, 1 mM E D T A buffer (50mM Tris buffer), homogenized at 4°C with a Dounce homogenizer and further disrupted by sonication for 2 minutes to release aldolase. The homogenate was centrifuged at 30 000 × g for 30 minutes at 4°C. The resulting supernatant was maintained at 4°C and used in subsequent assays of enzyme activity.
291 Aldolase assay Aldolase activity was assayed as described by Gracy, Lacko, Brox, Adelman and Horeoker 13 with the exception that 15/~g of glycerol phosphate dehydrogenasetriose phosphate isomerase mixture (Boehringer, Mannheim Corp.) was substituted for 10#g of each of the individual enzymes. One unit of aldolase activity is defined as the amount of enzyme catalyzing the cleavage of 1 #mole of fructose 1.6 diphosphate per minute at 25°C. Protein determinations Protein determinations were performed according to the method of Lowry, Rosebrough, Farr and Randall 14. Verification of cell normality To ascertain that no malignant transformation had occurred during in vitro passaging of the mouse fibroblasts, these cells were reinjected into isologous 2 month old recipients. Each recipient was injected intramuscularly with 4 × 106 cells from either the first passage or seventh passage (stationary phase). Each recipient mouse was examined for palpable tumors every fortnight for three months with no evidence of tumor growth in any of the mice injected. The diploid chromosome number was verified in cells of each passage as described by Merchant, Kahn and MurphylL A single exception to the recommended protocol was that cultures were exposed to colchicine for 16 rather than 8 hours. All preparations provided evidence that at each in vitro passage, the fibroblasts bore a normal diploid number of chromosomes per cell. Antisera to aldolase Monospecific antisera to purified mouse liver and mouse muscle aldolase from 3-4 month old mice were prepared as previously described 9,1°. The amount of aldolase antigen in each tissue culture homogenate was assayed by mixing a fixed amount o f homogenate (200/zl) with varying volumes of antiserum plus 50mM Tris buffer to make 250 #1 incubation mixture and incubated at 4°C overnight and then centrifuged at 3000 × g for 30 minutes at 4°C. The precipitated antigen-antibody complex was discarded and 100/zl of the supernatant was assayed for residual enzyme activity. Controls of both buffer and normal rabbit serum were always included. Serum alone showed low aldolase activity and this was subtracted from the recorded aldolase activity. RESULTS The growth parameters of the BALB/c mouse embryo fibroblasts in tissue culture can be seen in Fig. 1. By the sixth passage these cells have greatly slowed down in their ability to replicate and by the seventh transfer, they have virtually stopped growing. Preliminary experiments determined that the aldolase in the fibroblast cultures
292 Passages
I
-e--A--
]I 1I
5 -
--D--
E "Vr
2
{0
$ O
Z
I l
I 2
3
I 4
I 5
Days in culture
Fig. 1. Growth of mouse embryo fibroblasts as a function of passage in vitro. Cells were harvested by trypsinization and viable cell counts were performed with the aid of 0.05 ~ trypan blue in phosphate buffered saline.
is aldolase A (muscle type). Monospecific rabbit anti-mouse aldolase A both inactivated and precipitated this enzyme whereas monospecific rabbit anti-mouse aldolase B (river type) had virtually no effect on the aldolase activity of the culture homogenates. Determinations of enzyme activity per day in culture per passage showed that the specific activity of aldolase per mg of protein does not vary significantly from passage to passage, as seen in Fig. 2. Because the protein content of these
293 Passages
-a-
!
--&--
11'
-o-
11"
--II--
~"
--13-
"~
Z2
64
5.6
4.8 A
c
n o) "~E3.2 _ c !:3 "~ 2.4
~
1.6
I
I !
I 2
I 3
I 4
I 5
D a y s in c u l t u r e
Fig. 2. The specific activity of Aldolase A in the homogenates of individual cultures as function of time in culture. cultures decreased in proportion to cell number, this constant specific activity reflects a constant amount of aldolase activity per cell in each passage. A constant specific activity does not necessarily represent a constant level o f enzyme. There may well be inactive molecules (or molecules with reduced specific activity) present during different phases o f growth in vitro. In order to compare the amount o f aldolase antigen in each passage, cultures were harvested for the titration of enzyme by antibody three to four days after the initiation of each passage. Fig. 3 is a typical precipitation curve. In order to take advantage of the linear section of the precipitation curve, the quantity of enzyme precipitated by 1 F1 of antiserum
294
eQ
1.c e,-
~..~point of 50z precipitation
. i
Qe i
Io
0.2
I
I 2
I 4
I 6
I 8
I 10
Quantity of Antiserum (,ul)
I 12
Fig. 3. The precipitation of Aldolase A by antibody: A typical titration. The following calculation of units of enzyme precipitated by 1 #1 of antibody at the point of 50~ precipitation was made: Units of enzyme precipitated by 1/~l antiserum = units of enzyme precipitated at 50 ~ precipitation/ microliters of antisera required to precipitate 50 ~ activity was calculated from the point of 50 ~ precipitation. This calculation of the amount of enzyme precipitated by 1 #1 of antiserum with antigen in excess is valid over a wide range of initial enzyme activities, as shown in Fig. 4, and is, therefore, valid for comparing enzyme preparations which may contain sequestered enzyme present in the form of CRM. Comparison of the amount of enzyme activity precipitated by 1/A of antiserum on day 3 or 4 of each passage shows that the amount of enzyme antigen remains constant throughout the in vitro lifespan of mouse embryo fibroblasts (Fig. 5). Even though the specific enzyme activity per number of cells remains at a constant level, and there is no detectable evidence of an altered amount of enzyme antigen per successive passage in vitro certain minor alterations in enzyme structure might nevertheless occur which, under normal physiological conditions, might not express themselves in either altered enzyme activity or conformation. Such alterations might, however, express themselves under denaturing conditions e.g. heat inactivation. Homogenates of 3 or 4 day old cultures from each passage were, therefore, subjected to 60°C in a constant temperature water bath and the residual enzyme activity was assayed after various times of exposure. No differences in temperature sensitivity were observed over the entire #7 vitro life span of these fibroblasts. Figure 6 depicts the heat inactivation of enzyme from cells of passages I (secondaries) and VI. Enzyme from all intermediary passages showed similar heat sensitivity. Such a non linear (lst order) pattern of inactivation may reflect either a molecular heterogeneity
295
E "5
5"" Q4q
~
~..t•
,_,.,..gp•
"~Q24
"~
I
O
I
03
Initial
1,6
Aldolase
I
2.4
Activity (milliunil"s)
Fig. 4. The amount of enzyme activity precipitated by 1/d of antibody at 50% precipitation as a function of initial enzyme activity. Three secondary cultures were harvested and pooled before disruption. The resulting homogenate was assayed for aldolase activity and serial dilutions were made in 50 mM Tris Buffer to give aliquots with different initial aldolase activities. Each aliquot was titrated with antibody and the amount of enzyme precipitated by 1 /d of antiserum at 50% precipitation was performed as described in Fig. 3.
o f the aldolase which does not vary t h r o u g h o u t the in vitro lifespan, or selective heat sensitivity o f certain sites on the aldolase A molecule, while others are more resistant. DISCUSSION BALB/c mouse e m b r y o fibroblast cultures undergo a limited n u m b e r o f population doublings in vitro and then cease to replicate. In this respect mouse embryonic fibroblasts resemble fibroblasts o f other animal speciesZ,4, al. The p r o p o r t i o n o f cells capable o f synthesizing D N A in the mouse cultures decreases from 95 % in the first passage to 4 0 % in the last passage before the onset o f mitotic quiescence16; these results are similar to those reported for h u m a n fibroblasts whose life-span in vitro is limited17, a8 to about 50-70 doublings. We are aware that mouse embryo cells have been reported to show a high degree o f " s p o n t a n e o u s " neoplastic transformation when cultivated in vitro. However, when one scrutinizes carefully the data presented in papers reporting this p h e n o m e n o n it becomes obvious that this transformation and the subsequent f o r m a t i o n o f cell lines takes a few months to appear 19-z2. It is usually detected after an initial growth phase which is followed by a long quiescent period. The renewed growth after such a prolonged quiescent period is most likely due to a selection o f a small n u m b e r o f
296 E 2 E
.D ~(~0.64
0
~o.,~ t~
E N>'0.32 ¢...
u) 0.16 .~_ (.I
I
-n-
1
I
~I~ -I1Passages in culture
I
-7
I
I
Fig. 5. The amount of enzyme activity precipitated by 1 td of antiserum at 50% precipitation as a function of passage in culture. Cultures of passages I-III were assayed on day 3. Cultures of passages IV-VI were assayed on day 4.
transformed cells. It is unclear whether these cells appear in minute numbers during the growth phase or in the subsequent quiescent phase. Certainly growth conditions such as serum type 19-22 and cell density za exert a profound effect on the rate of transformation. There are several causes which one might suggest to explain the occurrence of this phenomenon of a limited life span of fibroblasts in vitro. One such suggested cause might be the production of altered enzyme molecules in the course of in vitro replications. Such enzyme molecules might result in an "error catostrophe" such as that suggested by Orgel z4,25 and, therefore, result in the death of the cultures. Alternatively, altered protein molecules may have a less drastic effect which could incapacitate the cell sufficiently to slow down its growth rate and eventually cause it to cease replication. Our findings of similar heat stability and a constant ratio of enzyme activity to enzyme antigen do not support the presence of altered enzyme molecules in mouse embryo fibroblast cultures which, nevertheless, posses a constant, limited in vitro lifespan. These findings are in contrast to parallel in vivo systems which have been reported from our laboratory 7-10 for in vivo experiments carried out both on nematodes and on mice which clearly reveal the accumulation of inactive enzyme molecules as a function of increasing age. Our results differ from those reported for human embryonic fibroblasts of the MRC-5 strain in which an accumulation of large amounts of altered protein leading to an "error catastrophe" has been reported. Lewis and Tarrant z~ have reported the
297
+I 8O
,I-
.5
I0
15
20
25
30
Time (minutes)
Fig. 6. The thermobility of Aldolase A from fibroblasts of in vitro passages I and VI. Culture homogenates were exposed to 60 °C in a constant temperature water bath. Aliquots were removed for assay of aldolase activity at the time intervals denoted above.
presence o f C R M to the enzyme, lactic dehydrogenase. This enzyme has 5 isozymes and even t h o u g h it has been reported that the relative amounts o f each isozyme remain constant t h r o u g h o u t the cultured lifespan o f these cells, interpretation o f such results is very difficult, especially when one is dealing with an apparently exceedingly abrupt decrease in detectable enzyme activity. Holliday and Tarrant 27 have studied the heat sensitivity o f the enzyme glucose-6-phosphate dehydrogenase (G-6-PD) and report an increase o f 20-25 ~ heat labile enzyme in late passage fibro-
298 blasts. In another publication, Holliday zs presents evidence that these same fibroblasts in culture undergo a selective process in which cells producing a new form o f G - 6 - P D become more and more frequent in culture. The selective advantage o f cells bearing a new isozymic form o f G - 6 - P D with, in all probability, a different sensitivity to heat denaturation might well account for the changes in heat stability observed in mass cultures. In contrast to the above findings, however, it should be noted that Martin and N o r w o o d (personal communication, T. N o r w o o d ) have not been able to detect any age-dependent C R M accumulation o f G-6-PD in WI-38 h u m a n fibroblasts in culture. Also, both Holland et al. 29 and Tomkins et al. 3° have employed viral probes for the detection of mis-synthesis o f proteins in h u m a n fibroblasts as a function o f age in culture. They showed that cells from early and terminal passages in culture were equally susceptible to herpesvirus type I, poliovirus type 129 and vesicular stomatitis virus30; the virus produced in cells o f all passages had the same specific activity and the senescent cells did not produce a larger p r o p o r t i o n o f defective virions. Holland et al. 29 also showed that viruses produced in " y o u n g " and "senescent" cultures have similar thermal stabilities and exhibit similar mutation rates. These results and the results reported in this communication indicate that whatever the cause o f the mortality o f fibroblasts in vitro it does not appear to be due to an error catasstrophe, nor, for that matter, to any general modification o f proteins. Relevant to this may be the recent (though as yet unconfirmed) findings o f Packer and Smith 31 which are consistent with the hypothesis that senescence o f fibroblasts in culture m a y be due primarily to oxidative damage. ACKNOWLEDGEMENTS This research has been supported by a grant from The Deutsche Forschungsgemeinschaft. We thank Mrs. Leah Rosenfelder for her skillful technical assistance.
REFERENCES 1 L. Hayflick and P. S. Moorhead, The serial cultivation of human diploid cell strains, Exp. Cell Res., 25 (1961) 585-621. 2 L. Hayflick, The limited in vitro lifetime of human diploid cell strains, Exp. Cell Res., 37 (1965) 614-636. 3 A. H. Ebeling, The permanent life of connective tssue outside of the organism, J. Exp. Med.. 17 (1913) 273-285. 4 G. M. Martin, C. A. Sprague and C. J. Epstein, Replicative life span of cultivated human ceils. Effects of donor's age, t!ssue and gcnotype, Lab. Invest., 23 (1970) 86-92. 5 L. Hayflick, The biology of human aging, Am. J. Med. Sci., 265 (1973) 432445. 6 H. Gershon, P. Zeelon and D. Gershon, Faulty proteins: Altered gene products in senescent cells and organisms. In A. Kohn and A. Shatkay (Eds.), ControlofGene Expression, Plenum, New York, 1974, pp. 255-264. 7 H. Gershon and D. Gershon, Detection of inactive enzyme molecules in aging organisms, Nature (London), 227 (1970) 1214-1217. 8 P. Zeelon, H. Gershon and D. Gershon, Inactive enzyme molecules in aging organisms: Nematode fructose-l,6-diphosphate aldolase, Biochemistry, 12 (1973) 1743-1750.
299 9 H. Gershon and D. Gershon, Altered enzyme molecules in senescent organisms: mouse muscle aldolase, Mech. Ageing Dev., 2 (1973) 3 3 4 1 . 10 H. Gershon and D. Gershon, Inactive enzyme molecules in aging mice: Liver aldolase, Proc. Natl. Acad. Sci. U.S.A., 70 (1973) 909-913. 11 L. Lima and A. Maciera-Coelho, Parameters of aging in chicken embryo fibroblasts cultivated in vitro, Exp. Cell Res., 70 (1972) 279-284. 12 R. J. Hay, R. A. Menzies, H. P. Morgan and B. L. Strehler, The division potential of cells in continous growth as compared to cell subcultivated after maintenance in stationary phase, Exp. Gerontol., 3 (1968) 3 5 4 4 . 13 R. W. Gracy, A. G. Lacko, L. W. Brox, R. C. Adelman and B. L. Horecker, Structural relations in aldolases purified from rat liver and muscle and Novikoff hepatoma, Arch. Biochem. Biophys., 136 (1970) 480-490. 14 O. H. Lowry, N. J. Rosebrough, A. L. Farr and R. T. Randall, Protein measurements with the Folin phenol reagent, J. Biol. Chem., 193 (1951) 265-275. 15 D. J. Merchant, R. H. Kalm and W. H. Murphy, Handbook o f Cell and Organ Culture, 2nd Ed. Burgess Publishing Co., Minneapolis, Minn., 1964. 16 M. Danot, Fetal mouse fibroblast cell cultures, a possible model for the study of aging? M.Sc. Thesis, Technion-Israel Institute of Technology, 1974. 17 A. Macieira-Coelho, Aging and Cell division, Front. Matrix, BioL, 1 (1973) 46-77. 18 V. J. Cristofalo and B. B. Sharf, Cellular senescence and D N A synthesis: thymidine incorporation as a measure of population age in h u m a n diploid cells, Exp. Cell. Res., 76 (1973) 419423. 19 K. K. Sanford, "Spontaneous" neoplastic transformation of cells in vitro: Some facts and theories, Nat. Cancer Inst. Monograph, 26 (1967) 387418. 20 K. K. Sanford, S. L. Handleman, J. W. Hartley, J. L. Jackson and R. R. Gnatt, Spontaneous neoplastic transformation in vitro: Influence of Endogenous murine leukemia virus and serum fractions, J. Natl. Cancer Inst., 49 (1972) 1177-1182. 21 V. J. Evans and W. F. Andersen, Effect of serum on spontaneous neoplastic transformations in vitro, J. Natl. Cancer lnst., 37 (1966) 247-249. 22 G. Carbone, R. Piazza and G. Parmiani, Effect of different sera on grwoth and "spontaneous" neoplastic transformation of mouse fibroblasts in vitro, J. Natl. Cancer Inst., 52 (1974) 387-393. 23 S. A. Aaronson, J. W. Hartely and G. J. Todano, Mouse leukemia virus : "Spontaneous" release by mouse embryo cells after long-term in vitro cultivation, Proc. Natl. Acad. Sci. U.S.A., 64 (1969) 87-94. 24 L. E. Orgel, The maintenance of the accuracy of protein synthesis and its relevance to aging, Proc. Natl. Acad. Sci. U.S.A., 49 (1963) 517-521. 25 L. E. Orgel, The maintenance of the accuracy of protein synthesis and its relevance to aging, Proc. Natl. Acad. Sci. U.S.A., 67 (1970) 1476. 26 C. M. Lewis, and G. M. Tarrant, Error theory and aging in human diploid fibroblasts, Nature (London), 239 (1972) 316-318. 27 R. Holliday and G. M. Tarrant, Altered enzymes in aging h u m a n fibroblasts, Nature (London), 238 (1972) 26-30. 28 R. Holliday, Humangenetik, 16 (1972) 83. 29 J. J. Holland, D. Kohne and M. V. Doyle, Analysis of virus replication in aging human fibroblast cultures, Nature (London), 245 (1973) 316-317. 30 G. A. Tomkins, E. J. Stanbridge and L. Hayflick, Viral probes of aging in the human diploid cell strain WI 38, Proc. Soc. Exp. Biol. Med., 146 (1974) 385-390. 31 L. Packer and J. R. Smith, Extension of the lifespan of cultured normal h u m a n diploid cells by vitamin E, Proc. Natl. Acad. Sci. U.S.A., 71 (1974) 4763-4767.
