Mechanisms o f Ageing and Development, 6 (1977) 185-195 185 ©Elsevier Sequoia S.A., Lausanne - Printed in the Netherlands

M A C R O M O L E C U L A R SYNTHESIS IN HUMAN F I B R O B L A S T S A T 37 ° AND 42 °C D U R I N G A G I N G I N V I T R O *

SAMUEL GOLDSTEIN and GREGORY W. CSULLOG

Departments of Medicine and Biochemistry, McMaster University Medical Centre, Hamilton, Ontario {Canada)

(Received September 8, 1976;in revised form November 4, 1976)

SUMMARY

A normal strain of human skin fibroblasts was pulse labelled at 42 ° and 37°C during inter~,als of the. in vitro lifespan with 3H-thymidine, aH-uridine and 14C-phenylalanine, precursors of DNA, RNA and protein synthesis, respectively. No increased heat lability could be detected in the relative profiles of precursor incorporation at the two temper- atures until terminal stages of passage when a minor increase in heat lability appeared in RNA and protein synthesis. Early passage cells, grown in the presence of the substrate analogues, p-fluorophenylalanine or 5-fluorouracil, showed a moderate increase in relative heat lability but comparable patterns were not evident in drug-free cultures, even at terminal stages of passage. Although the results indicate that defects exist in enzyme pathways for macromolecular synthesis at terminal stages of growth and in the presence

of analogues, they do not support the hypothesis that such defects are causal to the onset of cellular aging in vitro.

INTRODUCTION

Cultured human fibroblasts have a limited replicative capacity in vitro [1]. Although the underlying mechanism for this proliferative failure remains obscure, two general theories have been proposed: (1) an error theory, such that the synthesis of defective macromolecules eventually exceeds a lethal threshold [2]; (2) a genetic program, not unlike differentiation, wherein ceils lose genetic information needed to sustain proliferation [3]. Recent studies have shown that late-passage fibroblasts accumulate an increased proportion of heat-labile enzymes and alterations in other complex proteins [4-6]. Moreover, fibroblasts maintained at elevated temperatures have shortened lifespans, perhaps, as suggested, the result of increased infidelity in protein

*Presented at the 26th Annual Meeting of the Tissue Culture Association, June 4, 1975, Montreal, and in abstract form: In Vitro, 10 (1974) 381.

186

synthesis [7]. Other work on a cell mutant of the Chinese hamster ovary line has shown that the incorporation of labelled amino acids into protein becomes heat sensitive consequent to a structural alteration in leucyl tRNA synthetase [8].

In the present work, we have carried out studies on a strain of normal human fibroblasts during their in vitro lifespan, seeking heat-labile lesions in pathways leading to

synthesis of the three major macromolecules. The rationale was that if errors were

introduced into one or more enzymes involved in macromolecular synthesis, these might

be detectable as increased heat sensitivity in the rate of incorporation of labelled

precursors into DNA, RNA or protein. Fibroblasts were studied sequentially at 4 stages of the replicative lifespan, in some cases following growth in the presence of substrate

analogues known to lead to faulty protein synthesis and to induce premature senescence

[4, 51 .

MATERIALS AND METHODS

Cell culture

The cell strain used throughout these studies (DS) was derived from skin of the anterior forearm of a normal adult male, aged 36 years [6]. The maximum replicative potential was 60-65 mean population doublings (MPD).

Cells were routinely cultured in Eagle's minimum essential medium supplemented

with non-essential amino acids, glucose, pyruvate and 15% fetal calf serum (Gibco, Grand Island, N.Y.) and incubated in a humidified 95% air/5% CO2 atmosphere at 37 °C [9]. Subcultivation was carried out following trypsinization as described before [6, 9].

Studies were performed on a mass culture of cells reconstituted from liquid nitrogen storage at 4 MPD, with reference to the stage of harvesting explants (Fig. 1). At 9 MPD, some of the replicate dishes were grown continuously in the presence of 5- fluorouracil (5-FU) or p-fiuorophenylalanine (PFPA) while drug-free cultures were

maintained as controls. The first pulses were carried out when each culture reached 20 MPD under the

three conditions (Fig.l): drug-free control culture (pulse 1), PFPA-containing culture (pulse 1A), and 5-FU-containing culture (pulse 1B). Analogues, at the concentrations indicated below slowed the growth rates of the cultures so that more calendar time

elapsed to achieve the same level of MPD (Fig. 1). Pulses 2 and 3 were carried out at early and late "middle age", 36 and 50 MPD respectively. Growth of cells at 20 and 35 MPD

was similar, while cells at 50 MPD grew more slowly and to lower density, signs of early senescence [1, 4-6, 9]. More objective tests of senescent decline such as 3H-thymidine labelling index [10] were not done. Pulse number 5 (not shown in Fig. 1) was performed on a culture initiated from a second storage ampule of cells banked at late-passage and

further subcultured to 56 MPD.

Chemicals and isotopes DL-p-fluorophenylalanine and 5-fluorouracil were obtained from Calbiochem, La

Jolla, California. 3H-Thymidine (methyl-all, 6.7 Ci/mmol), 3H-uridine (6-all, 22.2 Ci/

187

¢C w

u J U > 4

,.I ~J n,'

T I M E I N D A Y S

1 9 33 50 58 91 135 156

O m

0 Z Z - < u,~

< z I-

0 10 20 30 40 50 60 70

M E A N P O P U L A T I O N D O U B L I N G S

Fig. 1. Scheme to measure macromolecular synthesis at 37° and 42 °C at different stages of the in vitro lifespan of cultured human fibroblasts (see Methods for further details).

mmol), and X*C-phenylalanine (uniformly labelled, 418 mCi/mmol) were obtained from

New England Nuclear Co. (Lachine, Quebec).

Preparation of cultures for pulse labelling When drug-treated and control cultures reached confluence at 20 MPD, cells were

harvested, transferred to 25 ml Erlenmeyer flasks at 1:4 splits and covered with loose-

fitting aluminum foil to allow free gas exchange. Cultures were incubated Overnight at

37 °C in the usual growth medium with and without drugs.

Pulse labelling with precursors of DNA, RNA and protein Procedures were modified from those of Thompson et al. [8]. Replicate flasks were

sealed with rubber stoppers and transferred to waterbaths at 37 ° or 42 °C and shaken through an amplitude of 6 cm at 50 cycles/min. After 0, 2, 4, 8, 12 or 24 hours of incubation at the two temperatures, the growth medium was aspirated and 2.0 ml of fresh, drug-free medium containing 0.5 /aCi/ml 3H-thymidine, or 0.5/aCi/ml 3H-uridine and 0.5 pCi/ml 14C-phenylalanine (the last two isotopes as a dual label) was added to

three flasks for each of the two combinations. After 30 minutes, flasks were removed from the baths, rinsed with 5 ml of ice-cold phosphate-buffered saline containing I00

ttg/ml each of non-radioactive thymidine, uridine and phenylalanine, and then treated with 5 ml of 5% trichloroacetic acid (TCA). After storage at 4 °C overnight, cell sheets

were detached by scraping with rubber policemen and the suspension was collected by f'dtering onto Whatman GF/C glass fibre filter discs. Discs were washed with three 10 ml volumes of ice-cold 5% TCA and one 5 ml volume of a 1:1 mixture (v/v) of diethyl ether

and 95% ethanol. They were allowed to air dry before being transferred to glass scintil- lation vials. 0.5 ml of hydroxide of hyamine solution (1 M in methanol) was added to

188

each vial and samples kept at 37 °C for three hours. Finally, 10 ml of scintillation fluor

(consisting of 0.05 g POPOP and 4.0 g PPO in 1 litre of toluene) was added to each vial,

the samples shaken and radioactivity determined on a Beckman liquid scintillation

spectrophotometer. Crossover of 14C into the 3H channel in the dual label samples and

quenching were corrected as before [11]. As a rule, zero time DPM, which were several

hundred-fold above background levels in early passage cells, fell during the replicative

lifespan but were always more than ten-fold background at terminal stages of passage. Data points represent the average of three determinations, except in a few cases where

some replicates were lost by contamination. Standard errors were generally within -+10%,

but occasionally -+15-20% for TdR incorporation. To compare results at the two temper-

atures, the incorporation of each isotope was corrected for the number of cells at the

corresponding time point in cultures at 37°and 42 °C. Results were then expressed as the radioactivity incorporated per cell at 42°C as a percentage of the incorporation per cell

a t37 °C.

Cell counting Replicate flasks were prepared as for pulse labelling and incubated at each of the

two temperatures. At the times noted, triplicates were removed, cell sheets harvested with

trypsin, suspended in phosphate-buffered saline and counted on a Cytograph ® electronic

cell counter [11] (Biophysics Corporation, Mahopac, N.Y.). Cell count determinations

had standard errors generally less than -+10% of the mean value with most of the

measurements ranging less than -+5%.

100 8O

6O

P ~ 1c

o .o

PULSE 1 (2OMPD)

~TdR

8

4 3 i ' - _ ~ ~ ,

2 e . . . "....

"... "....

1 ",m 8 16 24

TIME (Hours)

10

S

o

6 -"

i

Fig. 2. Relative synthesis of DNA, RNA and protein at 37 ° and 42 °C in drug-free cultures at early passage (MPD 20). Curves marked 37 °C and 42 °C indicate cell counts at the two temperatures. Curves marked TdR, Urd and Phe indicate the relative incorporation at the two temperatures of 31-t- thymidine, aH-uridine and Z4C-phenylalanine, precursors of DNA, RNA and protein, respectively, into acid precipitable material.

189

RESULTS

Cell numbers remained constant at 42 °C after an initial 20% decline during the zero time pulse interval (Fig. 2). There was a further decline to 24 hours. At 37 °C, by contrast, after an eight hour lag, cells proliferated vigorously.

The relative incorporation of aH-thymidine fell quickly at the higher temperature with a decrease of about 80% by the second hour of incubation and a steady, progressive

fall to undetectable levels at 24 hours (Fig. 2). The significant statistical variance for TdR incorporation (see methods) plus the rapid shutoff of incorporation at 42 °C in early passage cells limited the value of this parameter in seeking correlations between the heat- lability of DNA synthetic pathways and the MPD level. A lower temperature may have proved more useful but logistic considerations precluded experiments at a third temperature.

The incorporation of aH-uridine at 42 °C fell throughout the incubation interval. At zero time, RNA synthesis was almost the same at the two temperatures, but further incubation at 42 °C led to a steady decline. By 24 hours, it had fallen to less than half of that in the culture at 37 °C.

The incorporation of 14C-phenylalanine at 42 °C fell immediately over the first pulse period to 80% of the 37 °C value (Fig. 2). It then recovered during the following 4 hours and subsequently decreased slowly.

Experiments carried out on drug-free cultures at middle stages of the replicative lifespan are shown in Fig. 3. At both 36 and 50 MPD, cell counts again increased in cultures maintained at 37 °C, while at 42 °C, following an initial decline, there was an apparent recovery to zero time values.

PULSE 2 (36 MPD)

8o~/',, / , .... \ p ~ o O O 60~,. ~ . . . . "- , ,Urd%-, .

(~J 40 ['~ . . . . . , ,

"" "o 1C - 3 7 ~ . 4

_ = , g - - ~ 42o .

8 1 " . . . . . _~ 3 "-~-"~'" "..

o,,O 2 ":,

1 1 I ''''"J 0 8 16 24 0

TIME (Hours)

PULSE 3(50 MPD)

"- _

I r ~ :: 42 ° A~ . . - - '.~ -

-. I I ". 8 16 24

x

.i. U)

uJ O

Fig. 3. Relative synthesis o f DNA, RNA and protein at 37 ° and 42 °C in drug-free cultures at middle passage. All curves are labelled as in Fig. 2.

190

The profiles of incorporation for all three precursors in pulses 2 and 3 were similar

to those of pulse 1. In cultures incubated at 42 °C, DNA synthesis fell quickly, RNA synthesis fell slowly, while protein synthesis, following a characteristic initial drop

(perhaps more severe at 36 MPD), first recovered to the level in cultures at 37 °C and then declined. The major consistent difference between cultures at the 3 passage levels was the higher relative incorporation of 14C-phenylalanine in the 8 hour pulse at 36 and 50 MPD.

However, with the possible exception of protein synthesis in pulse 2, no increases were observed in the heat-lability of synthetic pathways above those of pulse number 1.

PULSE 4 (61MPD)

1 0 0 .

°~ 20-...v.TdR k Ur d •

" i Z 10 g 8

~ e - ~ 5 - - n- 4 - "

" ' " ' . .

PULSE 5 (56MPD)

k ; x

m

w

I ~...""i 8 16 240

TIME (Hours)

Fig. 4. Relative synthesis of DNA, RNA and protein at 37 ° and 42 °C in drug-free cultures at late- passage. Curves are labelled as described in Fig. 2.

Cells at 61 MPD (Fig. 4) were clearly senescent and only 2 MPD from complete termination of growth. The population doubling time of the culture was 3-4 times that of the culture at MPD 20 (about 70 v e r s u s 20 h), cells were sparse and covered by substantial granular debris. As with previous pulses, the incorporation of 3H-thymidine at 42 °C fell quickly in the first 2 hours, and eventually, to undetectable levels. Some differences were observed in the profiles of RNA and protein synthesis. First, 3H-uridine incorporation at 42 °C showed an earlier fall, dropping to 80% of the 37 °C value in the zero hour pulse period, to about 45% at 4 h, followed by a recovery phase and finally, a decline. Secondly, the recovery phenomenon for ~4C-phenylalanine incorporation took longer to complete, reaching only 80% of the 37 °C value by the fourth hour of incubation. It also appeared to overshoot prior to the usual second phase of decline. However, the cell population decreased in the latter half of incubation at both 37 ° and 42 °C probably due to detachment and/or lysis of these debilitated cells. Therefore, DS cells from another ampule were reconstituted from liquid nitrogen at late middle-

191

passage, grown to MPD 56, and the pulse experiment repeated (Fig. 4). Cell counts at the two temperatures remained relatively constant indicating that although the proliferative

potential was low, cells were healthy enough to survive the conditions. Except for a

partial rebound of 3H-TdR incorporation at 24 h, and the absence of overshoot in protein

and RNA synthesis at 8 h, the sensitivity of this culture to the high temperature was

similar to that of cultures at early and middle passage (pulses 1-3).

To choose suitable concentrations of each analogue for pulse studies, dose-response

experiments were first carried out on cell growth. Growth was retarded progressively in

cultures maintained at concentrations of PFPA above 10-SM (Fig. 5). PFPA was about 50% inhibitory at 10-4M, while it completely inhibited cell proliferation at 5 × 10-4M.

9 ~CONTROL J / 10"SM 'x~v~'r-- 87 ~ lO'eM

~ s

10-4M 25 o.

~ , 4

~3

Z

1 ~ 5 x 1 0 " 4 M

I I I I I I 2 3 4 5 6 7 DAYS AFTER PLATING

Fig. 5. Effect of p-fluorophenylalanine on the proliferation of cultured human fibroblasts. 10 $ cells were seeded into 60 mm Petri dishes when sells were at 9 MPD. Analogue was present at the concen- trations shown and cultures were incubated under standard conditions. Replicate dishes were harvested for cell counts as described in Methods.

The effects of 5-FU on the proliferation of cells are summarized in Table I. Inhibition first appeared at 2.4 X lO-TM and progressed to 19.2 × lO-TM, the highest concentration, which was completely inhibitory to cell growth.

The concentration of each analogue partially inhibitory to growth was present in growth medium continuously from 9 MPD until cells were prepared for pulsing at 20

MPD (Fig. 6). In pulse 1A (PFPA-treated) the relative incorporation of both 3H- thymidine and 3H-uridine at zero times was lower than previously seen in drug-free cultures. In the 2-8 hour periods, profiles looked about the same as in pulse 1. Synthesis of all three macromolecules fell rapidly to undetectable levels between 8-24 h, but cell counts also fell to near zero at both temperatures. This indicates that PFPA plus the continuous shaking during incubation, operating together, produced cell detachment or

192

TABLE I

EFFECT OF 5-FLUOROURACIL ON THE PROLIF- ERATION OF HUMAN FIBROBLASTS

Analogue Time to reach concentration, confluency × 10 -7 {M) (days)

0.0 4 1.2 4 2.4 8 4.8 9 9.6 12

19.2 No growth

Cells were seeded into culture dishes at 1:8 splits and analogue was added at the concentrations shown. The time for cells to reach confluency, as determined by cell population densities and the absence of mitotic figures, was recorded for each culture.

PULSE 1A 10-4M PFPA (20 MPD)

10o 80 ',~Urd 1 Phe

6 o "-,

,~ ,o i \ =~ 3o ~T<i. \ u) i

o '.. ' • '¢- 1oi ": o 8 - s + f-i3,o~ \

o.o7 '"\ 1 I A'"" ,,

8 16 24

PULSE 1B 9.6 x 10-7M 5FU (20 MPD)

~ P h e

L -TdR ',,,,,, ,,,

- ",, 5

~ v o "'~'o 3 ~

I ) I ''" "":'o ° ~ 0 8 16 24

TIME (Hours)

Fig. 6. Relative synthesis of DNA, RNA and protein at 37 ° and 42 °C in analogue-containing cultures at MPD 20. Curves are labelled as described in Fig. 2.

lysis. Parallel cultures cultivated without shaking in the presence of the same PFPA

concentration at 37 °C did not detach, and in fact, were capable of continued slow

growth as in Fig. 5. Even though interpretation is obscured by this event, there appeared

to be increased heat-labilities in all three synthetic pathways compared to control cultures

up to 12 h, the last time point at which cells remained viable.

The culture maintained with 5-FU (pulse 1B) showed increased heat-lability for 31-1-

uridine and 14C-phenylalanine incorporation, but to a lesser degree than cultures treated

193

with PFPA. Despite the presence of 5-FU, cell counts actually increased, roughly in parallel at both temperatures.

DISCUSSION

In the present experiments we have been unable to find significantly increased heat- lability in the rate of incorporation of labelled precursors into DNA, RNA and protein of cultures that have begun to show early evidence of senescent growth decline. It must be emphasized that such changes may very well exist, but go undetected by the relatively crude methods employed here. A moderate degree of temperature sensitivity is discernible in RNA and protein synthesis at terminal stages, but this comes well beyond the onset of senescence in experiments where the original aim was to assign causality to error mechanisms. However, the results are consistent with previous data that late-passage fibroblasts contain an increased fraction of heat-labile enzymes [4-6], and not in accord with a report which fails to demonstrate these changes [12].

It is possible that one or more key enzymes within the three synthetic pathways / .

develop structural alterations at the onset of in vitro senescence, but these are not heat- labile. One form of the error hypothesis predicts [2] that consequent to defects in the key information-handling amino acid synthetases, a variety of amino acid mis-sense substitutions occur to produce a family of altered proteins; in terms of sensitivity to heat, some of these would be more labile, some more stable, and others unchanged. An alternate form of error hypothesis, based on multiple somatic mutations [2, 13] would produce a similar picture. In an3/case, our experiments, which measure the phenotypic resultant of many vectors, would miss a significant heat-sensitive lesion if its effects were cancelled by one or more of opposite heat sensitivity. In contrast, a single discrete mutation in Chinese hamster ovary cells, affecting leucyl amino acid synthetase, is associated with a profound and consistently increased heat-lability in the rate of incorporation of labelled amino acids into protein [8].

The immediate drop in the incorporation of 14C-phenylalanine at 42 °C followed by recovery to the incorporation levels at 37 °C is of interest and in accord with experiments on HeLa cells [14, 15]. These studies demonstrate that the recovery phenomenon at 42 °C is dependent upon RNA synthesis and may relate to a factor that stimulates the initiation o f protein synthesis. The present work suggests that this factor exists in human fibroblasts as well, although it is apparently not affected by aging in vitro.

Substrate analogues have been used to introduce errors into protein and RNA. Thus PFPA has been shown to be incorporated into bacterial and mammalian proteins in vivo

and in vitro [16]. Similarly, 5-FU is incorporated into RNA leading to changes in the properties of ribosomes, tRNA and mRNA [17, 18]. Indeed, 5-FU was shown to induce premature senescence in MRC-5, a strain of human fetal lung fibroblasts, and this was preceded by the appearance of heat-labile enzyme [4]. However, when WI-38 cells, also of fetal lung origin, were grown in the presence of the amino acid analogues PFPA and ethionine at concentrations that did not stop growth acutely or irreversibly, they

194

failed to undergo premature senescence [19]. In the present studies, concentrations of analogues that retarded cell growth did, in fact, lead to a more rapid fall of precursor incorporation at 42 °C. But similar heat sensitivities were not apparent in middle-passage cultures free of analogues and only to a minor degree in late-passage cells. We recognize, however, that other factors could contribute to differing profiles of precursor incorporation following analogue treatment and elevated temperature. Thus, it is possible that qualitative or quantitative alterations occur in membrane lipids or proteins. Perhaps as a consequence, there may be differential effects on the sizes, types and availability of precursor pools for macromolecular synthesis.

Alternate explanations exist for the .present negative results. It is possible that proteolytic mechanisms rapidly identify and degrade the putative faulty proteins that occur in untreated cultures at the onset of senescence. Thus, while incubation at elevated temperatures may promote the synthesis of defective proteins both directly, at the level of translation [7], and indirectly, by increasing the mutation frequency [20], it also enhances proteolysis [21]. Indeed, MRC-5 and WI-38 cells beyond middle-passage show accelerated proteolysis at 37 °C, perhaps in response to the presence of defective proteins

[22, 23]. On the other hand, WI-38 cells at late-passage are less able to eliminate analogue-containing proteins compared to cells at early passage [22] perhaps because proteolytic enzymes become defective themselves due to the analogues. This idea of defective proteolysis in late-passage cells is supported by other work demonstrating a decreased ability to degrade certain proteins in skin fibroblasts [11] as well as the loss of a neutral protease activity in late-passage WI-38 cells [24].

Evidence also exists for defects related to infidelity in DNA replication [25]. Thus, DNA polymerase of late-passage MRC-5 cells copies DNA with a significantly higher error frequency than that of young cells. Additionally, the capacity to repair DNA declines in late-passage fibroblasts [26, 27]. It seems clear, therefore, that one or more mechanisms could lead to a multiplicity of defects in old cells with widespread involvement of DNA, RNA and protein metabolism. However, none of the data so far available, including the present results, provide evidence for such lesions at the onset of senescence, and hence, do not support the idea of an increasing error cascade that is causal to aging of normal cells in vitro.

ACKNOWLEDGEMENTS

This work was supported by grants from the Medical Research Council of Canada during the tenure of an MRC Scholarship (S.G.) and Studentship (G.W.C.). We thank Elena Moerman and Shelley Ballantyne for excellent technical assistance.

REFERENCES

1 L. Hayflick, The limited in vitro lifetime of human diploid cell strains, Exp. CellRes., 37 (1965) 614-636.

195

2 L. E. Orgel, Aging of clones of mammalian cells, Nature, 243 (1973) 441-445. 3 B. L. Strehler, G. Hirsch, D. Gusseck, R. Johnson and M. Bick, Codon-restriction theory of aging

and development, J. Theor. Biol., 33 (1971) 429-474. 4 R. Holliday and G. M. Tarrant, Altered enzymes in ageing human fibroblasts, Nature, 238 (1972)

26-30. 5 C. M. Lewis and G. M. Tarrant, Error theory and ageing in human diploid fibroblasts, Nature, 239

(1972) 316-318. 6 S. Goldstein and E. J. Moerman, Defective proteins in normal and abnormal human fibroblasts

during aging in vitro, Interdiscip. Top. Gerontol., 10 (1976) 24-43. 7 K. V. A. Thompson and R. Holliday, Effect of temperature on the longevity of human fibroblasts

in culture, Exp. Cell Res., 80 (1973) 354-360. 8 L. H. Thompson, J. L. Harkins and C. P. Stanners, A mammalian cell mutant with a temperature-

sensitive leucyl-transfer RNA synthetase, Proc. Nat. Acad. Sci. USA, 70 (1973) 3094-3098. 9 S. Goldstein and J. W. Littlefield, Effect of insulin on the conversion of glucose-C-14 to C-14-O2

by normal and diabetic fibroblasts in culture, Diabetes, 18 (1969) 545-549. 10 V. Cristofalo and B. B. Sharf, Cellular senescence and DNA synthesis. Thymidine incorporation as

a measure of population age in human diploid cells, Exp. Cell R es., 76 (1973) 419-4 27. 11 S. Goldstein, D. Stotland and R. A. J. Cordeiro, Decreased proteolysis and increased amino acid

efflux in aging human fibroblasts, Mech. Ageing Dev., 5 (1976) 221-233. 12 W. R. Pendergrass, G. M. Martin and P. Bornstein, Evidence contrary to the protein error

hypothesis for in vitro senescence, J. Cell PhysioL, 87 (1976) 3-14. 13 S. J. Fulder and R. Holliday, A rapid rise in cell variants during the senescence of populations of

human fibroblasts, Cell, 6 (1975) 67-75. 14 W. McCormick and S. Penman, Regulation of protein synthesis in HeLa ceils: Translation at

elevated temperatures, J. Mol. Biol., 39 (1969) 315-333. 15 S. Penman, E. Goldstein, M. Reichman and R. Singer, The relation of protein synthesis regulation

and RNA metabolism, Acta Endocrinol. Suppl., 180.(1973) 168-191. 16 J. S. Fruton, in H. Neurath (ed.), The Proteins, Vol. 1, Academic Press, New York, 1963, pp.

189-310. 17 H. G. Mandel, The incorporation of 5-fluorouracil into RNA and its molecular consequences, Prog.

Mol. Subcell. Biol., 1 (1969) 82-135. 18 P. Roy-Burman, 5Fluorouracil, 5-Fluorouridine, and 5-Fluoro-2'-deoxyuridine, in Recent Results

in Cancer Research, Vol. 25, Springer-Verlag, New York, 1970, pp. 49-55. 19 J. M. Ryan, G. Duda and V. J. Cristofalo, Error accumulation and aging in human diploid cells, J.

Gerontol., 29 (1974) 616-621. 20 J. W. Drake and R. H. Baltz, The biochemistry of mutagenesis, Ann. Rev. Biochem., 45 (1976)

11-37. 21 A. L. Goldberg and A. C. St. John, Intracellular protein degradation in mammalian and bacterial

cells: Part 2,Ann. Rev. Biochem., 45 (1976) 747-803. 22 M. O. Bradley, L. Hayflick and R. T. Schimke, Protein degradation in human fibroblasts (WI-38),

Effects of aging, viral transformation, and amino acid analogs, J. Biol. Chem., 251 (1976) 3521-3529.

23 V. Shakespeare and J. H. Buchanan, Increased degradation rate of protein in aging human fibroblasts and in cells treated with an amino acid analog, Exp. Cell Res., 100 (1976) 1-8.

24 H. B. Bosmann, R. L. Gutheil and K. R. Case, Loss of a critical neutral protease in ageing WI-38 cells, Nature, 261 (1976)499-501.

25 S. Linn, M. Kairis and R. Holliday, Decreased fidelity of DNA polymerase activity isolated from aging human fibroblasts, Proc. Nat. Acad. ScL USA, 73 (1976) 2818-2822.

26 S. Goldstein, The role of DNA repair in aging of cultured fibroblasts from Xeroderma Pigmentosum and normals, Proc. Soc. Exp. Biol. Med., 137 (1971) 730-734.

27 J. B. Little, Relationship between DNA repair capacity and cellular aging, Gerontology, 22 (1976) 28-55.