Age-related telomere shortening occurs in lens epithelium from old rats and is slowed by caloric restriction.

We have investigated whether the average relative telomere length of lens epithelial cells (LECs) from brown Norway rats decreases with the age of the donor animal, and whether chronic caloric restriction (CR) of the rats delays the telomere shortening. Our previous studies have demonstrated that clonal proliferative potential of rodent LECs as well as the in vivo rate of DNA synthesis decreases with age and that this decrease is slowed by chronic lifelong caloric restriction (CR). In order to determine if telomeric shortening might be involved in this loss of proliferative potential, we examined relative telomeric lengths in young, old ad lib fed (AL), and old calorically restricted (CR) brown Norway rats. We used fluorescence in situ hybridization with a peptide nucleic acid probe (PNA) complementary to the telomeric repeat sequence to quantitate relative telomere lengths in LECs in lens sections (TELO-FISH). Control experiments demonstrated that the PNA probe binding was restricted almost entirely to the terminal portions of the rat chromosomes with less than 5% bound at interstitial sites in typical metaphase spreads. The relative telomere lengths of interphase human fibroblast standards, as determined by TELO-FISH, were in good agreement with terminal restriction fragment analyses of the same standards and with literature values for rat cells. The average telomere lengths of interphase nuclei in the old AL rat LECs were found to be 21% shorter than paired young AL controls (P < 0.01 by Wilcoxian signed rank test). The calorically restricted old rats had less telomere erosion (12%) than the old AL group (P < 0.05). Although it is not clear whether such moderate telomeric erosion can limit cell division in rodent LECs, the telomeric shortening correlated well with previous studies demonstrating reduced clonal, replicative potential, and reduced rates of in vivo DNA replication in LECs from old rodents and a delay in this attenuation in animals on chronic CR.

Cellular proliferation potential during aging and caloric restriction in rhesus monkeys (Macaca mulatta).

Caloric restriction (CR) is the most successful method of extending both median and maximal lifespans in rodents and other short-lived species. It is not yet clear whether this method of life extension will be successful in longer-lived species, possibly including humans; however, trials in rhesus monkeys are underway. We have examined the cellular proliferative potential of cells from CR and AL (ad libitum fed) monkey skin cells using two different bioassays: colony size analysis (CSA) of dermal fibroblasts isolated and cloned directly from the skin and beta-galactosidase staining at pH 6.0 (BG-6.0) of epidermal cells in frozen sections of skin. Decreases in both proliferative markers occurred with age, but no differences were observed between CR and AL animals. Skin biopsies were obtained from AL and CR rhesus monkeys from two different aging colonies, one at the National Institute on Aging (NIA) and one at the University of Maryland-Baltimore (UMB). These biopsies were used as a source of tissue sections and cells for two biomarkers of aging assays. The CR monkeys had been maintained for 9-12 years on approximately 70% of the caloric intake of control AL animals. In the CSA studies, the fraction of small clones increased significantly and the fraction of large clones decreased significantly with increasing age in AL monkeys. The frequency of epidermal BG-6.0 staining cells increased with age in older (>22 years) AL monkeys, but most predominately in those of the UMB colony, which were somewhat heavier than the NIH AL controls. Old monkeys on CR tended to have fewer BG-6.0-positive cells relative to old AL-derived epidermis, but this effect was not significant. These results indicate that cellular proliferative potential declined with age in Macaca mulatta, but was not significantly altered by CR under these conditions. Although these experiments are consistent with an absence of effect of CR on monkey skin cell proliferative potential, we have found in previous experiments with mice that a longer duration of CR (as a fraction of total lifespan) was needed to demonstrate CR-related improvement in clone size in mice. Further studies on the now mid-aged monkeys will be needed as their age exceeds 20 years to conclusively rule out an effect of CR on proliferative potential of skin cells from these primates.

Analysis of the capacity of extracts from normal human young and senescent fibroblasts to support DNA synthesis in vitro.

Cytoplasmic extracts from early-passage (young), late-passage (senescent) normal human fibroblast (HF) cultures and immortalized human cell lines (HeLa, HT-1080, and MANCA) were analyzed for their ability to support semiconservative DNA synthesis in an in vitro SV40-ori DNA replication system. Unsupplemented extracts from the three permanent cell lines were demonstrated to be active in this system; whereas young HF extracts were observed to be minimally active, and no activity could be detected in the senescent HF extracts. The activity of these extracts was compared after supplementation with three recombinant human replication factors: (1) the catalytic subunit of DNA polymerase alpha (DNA pol-alpha-cat), (2) the three subunits of replication protein A (RPA), and (3) DNA topoisomerase I (Topo I). The addition of all three recombinant proteins is required for optimum activity in the young and senescent HF extracts; the order of the level of activity is: transformed > young HF > senescent HF. Young HF extracts supplemented with RPA alone are able to support significant replicative activity but not senescent extracts which require both RPA and DNA pol-alpha-cat for any detectable activity. The necessary requirement for these factors is confirmed by the failure of unsupplemented young and senescent extracts to activate MANCA extracts that have been immunodepleted of DNA pol-alpha-cat or RPA. Immunocytochemical studies revealed that RPA, DNA pol-alpha, PCNA, and topo I levels are higher in the immortal cell types used in these studies. In the HF cells, levels of DNA pol-alpha-cat and PCNA are higher (per mg protein) in the low-passage than in the senescent cells. By contrast, RPA levels, as determined by immunocytochemical or Western blot studies, were observed to be similar in both young and senescent cell nuclei. Taken together, these results indicate that the low to undetectable activity of young HF extracts in this system is due mainly to reduced intracellular levels of RPA, while the senescent HF extracts are relatively deficient in DNA polymerase alpha and probably some other essential replication factors, as well as RPA. Moreover, the retention of RPA in the senescent HF nuclei contributes to the low level of this factor in the cytoplasmic extracts from these cells.

The relationships of animal age and caloric intake to cellular replication in vivo and in vitro: a review.

This brief review examines aging at the cellular level as expressed by cell replication rates in vivo, clone size limits in vitro, and cell function in several tissues and organs. Studies are presented in which in vivo and in vitro cell replication measurements were made for several cell types and organs in relation to animal age, diet, life span, and specific age-related pathologies. Among the events examined that affect cell replication and cell survival in vitro and in vivo over a lifetime are oxidative damage, telomere shortening, and hormone and hormone receptor level changes. Long-term caloric restriction (CR) is favorable or protective for all of these events when measured in later life and comparisons are made to ad libitum (AL)-fed animals, and it is accompanied by more youthful rates of cell replication. It is proposed that in vivo and in vitro measures of cellular replication constitute biomarkers of aging when applied to comparisons of CR and AL diet rodents, where they correlate with the delay of disease and extension of life span. Longitudinal studies are needed to confirm this. The occurrence of certain age-related pathophysiologic states, such as immune (T cell) insufficiency, cataract, and senile osteopenia/osteoporosis, are accompanied by major diminishments of replication rates, numbers, and functions of the essential cell types in the organs and tissues involved. However, direct evidence is lacking that diminished cell replication in specific organs contributes to the limitation of life span.

Response of lens epithelial cells to hydrogen peroxide stress and the protective effect of caloric restriction.

Hydrogen peroxide (H2O2) has been reported to be present at significant levels in the lens and aqueous humor in some cataract patients and suggested as a possible source of chronically inflicted damage to lens epithelial (LE) cells. We measured H2O2 effects on bovine and mouse LE cells and determined whether LE cells from old calorically restricted mice were more resistant to H2O2-induced cellular damage than those of same age ad libitum fed (AL) mice. Bovine lens epithelial cells were exposed to H2O2 at 40 or 400 microM for 2 h and then allowed to recover from the stress. The cells were assayed for DNA damage, DNA synthesis, cell viability, cell morphology, response to growth stimuli, and proliferation potential. Hydrogen peroxide-treated cells showed an increased DNA unwinding 50% greater than that for untreated controls. These DNA strand breaks appeared to be almost completely rejoined by 30 min following removal of the cells from a 2-h exposure. The 40 microM exposure did not produce a significantly lower DNA synthesis rate than the control, it responded to growth factor stimuli, and it replicated as did the control cells after removal of H2O2. The 400 microM H2O2 severely affected DNA synthesis and replication, as shown by increased cell size and by markedly reduced clonal cell growth. The cells did not respond to growth stimulation by serum or growth factors and lost irreversibly the capacity to proliferate. The responses of LE cells from old adlib diet (AL) and calorically restricted (CR) mice to H2O2 were significantly different. Exposure of LE cells to 20, 40, or 100 microM H2O2 for 1 h induces a significant loss of cellular proliferation in cells from old AL mice. LE cells from long-term CR mice of the same strain and age were more resistant to oxidative damage at all three concentrations of H2O2 than those of both old and young AL mice and showed a significantly higher proliferation potential following treatment. It is concluded that CR results in superior resistance to reactive oxygen radicals in the lens epithelium.

Enhanced cell proliferation and biosynthesis mediate improved wound repair in refed, caloric-restricted mice.

Aged mice that have undergone long-term caloric-restriction (CR) have improved health and enhanced longevity in comparison to aged mice that are ad libitum-fed (AL). However, caloric-restriction does not benefit the impaired wound healing of aged mice. To test the hypothesis that CR mice have the capacity for enhanced wound repair, but require a short-term period of additional nutrient intake to show this advantage, we assessed wound healing in CR mice that had been refed (RF) an ad libitum diet for 4 weeks prior to wounding. Two strains of AL young (Y AL) (4-6 months), AL middle-aged (M AL) (15-17 months), and three different, matched cohorts of old mice (O) (30-33 months): O AL, O CR, and O RF were studied. Two full-thickness 4 mm diameter punch biopsy skin wounds were created on the dorsum of each mouse. Animals were sacrificed and wounds were harvested at 1,2,3,5, and 7 days post-wounding. Repair of wounds was slower in O AL and O CR mice compared to Y AL and M AL animals. In contrast, the O RF mice healed similarly to that of the Y AL and M AL mice, as assessed by measures of wound area and histologic criteria. O RF mice demonstrated enhanced synthesis of type I collagen mRNA in comparison to O AL and O CR mice. A greater number of endothelial cells and fibroblasts at the wound edge of the O RF mice exhibited replication in vivo as measured by uptake of BrdU. O RF mice had higher levels of insulin-like binding protein 3 (IGFBP-3). Furthermore, fibroblasts derived from the explant of the punch biopsy of O CR mouse skin revealed enhanced proliferation and contraction in vitro, in comparison to fibroblasts from the O AL mice. In conclusion, O RF mice demonstrate an enhanced capacity to undergo wound repair in comparison to O AL mice. This effect appears to be mediated, in part, by enhanced cell proliferation, contraction, and collagen biosynthesis. In addition, short-term refeeding induced an increase in the serum level of IGFBP-3, the major binding protein for IGF-1. These data confirm that cells from O CR animals have a preserved proliferative, biosynthetic, and contractile capacity, but that an adequate source of nutrients is necessary to demonstrate this advantage in wound healing.

Caloric restriction: conservation of cellular replicative capacity in vitro accompanies life-span extension in mice.

We have tested whether life-long caloric restriction (CR) slows or delays the age-related loss of cellular replicative potential that occurs during normal aging in ad libitum (AL) fed mice. Both mean and maximum life spans of the restricted animals (60% of AL intake) were significantly extended 30-40% by CR treatment. Proliferative potential, measured by determining the fraction of cells capable of forming large clones in vitro, was compared in five cell types from six tissue sites from two strains of mice (Male (C57BL/6 x DBA/2)F1("B6D2F1") and female (C57BL/6 x C3H)F1("B6C3F1")). This included four nonhematopoietic organ sites: fibroblast cells from ear skin, tail skin, and subdermal connective tissue and epithelial cells from the medullary part of the kidney and two cell types, myofibroblasts and endothelial-like cells, from spleen and bone marrow. The proliferative potential of cells from AL mice decreased progressively with age in all tissues sites of both mouse strains. CR delayed or decreased the loss of proliferative potential in all situations, but the timing of this was tissue specific. For cells from the four nonhematopoietic tissues sites from female B6C3F1 female mice, CR delayed the onset of proliferative loss, such that the fraction of large clones was significantly greater for the CR 18- to 24-month-old mice than in AL controls at three of four sites (as determined by the fraction of large clones after 1 week of clonal growth). The proliferative loss in CR tissues then accelerated from 24 to 30 months, so that both CR and AL mice had similar fractions of large clones after 30 months of age. CR was also seen to delay loss of proliferative potential in cells from skin and kidney of B6D2F1 male mice at 23-24 months of age when cloned for 2 weeks. For fibroblast and endothelial-like cells from bone marrow and spleen stromal sites from both strains of mice, CR also significantly decreased loss of proliferative potential; furthermore, in these tissues the proliferative advantages remained or increased from 24 to over 30 months of age. In companion studies (N.S. Wolf et al., 1995. Exp. Cell. Res. 217, 000-000), CR was seen to decrease age-related losses in the maximal rates of cell replication in vivo in a panel of tissues from B6D2F1 male mice.(ABSTRACT TRUNCATED AT 400 WORDS)

Caloric restriction: conservation of in vivo cellular replicative capacity accompanies life-span extension in mice.

In male mice of a long-lived hybrid strain (B6D2F1), long-term 40% caloric restriction (CR) extended both mean and maximum life spans by 36 and 20%, respectively, over that of ad libitum fed (AL) controls. Measurements of entry into S-phase were made in vivo of six different cell types in five different organs using 2-week exposures to BrdU. The labeling index (L.I.) in all organs studied was lower in young CR mice than in young AL fed mice. In most cases, the L.I. in AL mice fell to the levels of that in the CR mice by 13 months of age, and the two groups then remained so through old age. However, when the L.I. was measured in old CR mice which had been placed on the AL diet for a period of 4 weeks (this was termed refeeding (RF), it was found to be above that of similar age AL or CR mice and almost at the level of young AL mice. This was still true, but to a lesser degree, in a repeat study using an 8-week period of RF. In a separate but parallel in vitro study (companion paper, this volume), the superiority of CR over AL for retention of cellular replication capacity was confirmed by clone size distribution measurements made in several cell types in mice of several age groups. These results indicate that: (1) the rate of cell replication in AL diet mice diminishes greatly by early middle age in all organ sites studied and then plateaus or declines much more slowly; (2) CR broadly preserves in vivo cellular replicative capacity but often requires the energy levels provided by a switch to AL feeding to demonstrate this late in life; (3) accordingly, the replicative deficit in AL fed mice appears to be cumulative and is significant only in old age. The mechanism(s) involved is yet to be discovered but may be related to, or even the same as, that which extends life spans in CR animals. Correspondingly, and with corroborative data from our in vitro companion study, (W. R. Pendergrass et al., 1995. Exp. Cell. Res. 217, 309-316), we suggest that cell populations sustain an accrual of biochemical damage or physiological alterations which increasingly limit their replicative capacity as the animal ages, and that CR reduces the accrual of this damage.

Murine temperature-sensitive DNA polymerase alpha mutant displays a diminished capacity to stimulate DNA synthesis in senescent human fibroblast nuclei in heterokaryons at the nonpermissive condition.

We have investigated the capacity of a murine cell line with a temperature-sensitive (ts) mutation in the DNA polymerase alpha (Pola) locus and a series of ts non-Pola mutant cell lines from separate complementation groups to stimulate DNA synthesis, in senescent fibroblast nuclei in heterokaryons. In the Pola mutant x senescent heterodikaryons, both human and murine nuclei display significantly diminished levels of DNA synthesis at the restrictive temperature (39.5 degrees C) as determined by [3H]thymidine labeling in autoradiographs. In contrast, all of the non-Pola mutants, as well as the parental (wild type) murine cells, induced similar levels of DNA synthesis in both parental nuclei at the nonpermissive and permissive temperatures. Similarly, young human fibroblasts are also able to initiate DNA synthesis in heterokaryons with the ts Pola mutant at the two temperatures. In order to determine if complementation of the non-Pola mutants requires induction of serum responsive factors in the senescent cells, fusion studies of similar design were conducted with young and old human fibroblasts incubated in low serum (0.2%) for 48 hr prior to and after cell fusion. Again, a diminished level of DNA synthesis was observed at 39.5 degrees C in the Pola mutant x senescent cell heterokaryons. In these low-serum studies, both parental nuclei in the Pola x young cell heterokaryons and the human nuclei in heterokaryons with one of the non-Pola mutants (FT107) also displayed diminished levels of DNA synthetic activity. All of the other mutants are able to support similar levels of synthetic activity at both temperatures in the presence of reduced serum. The nature of the mutation in three of the non-Pola lines has not been determined but, like the Pola mutant cells, are inhibited in the G1 phase of the cell cycle when incubated at the nonpermissive temperature (39.5 degrees C). The fourth non-Pola mutant line is known to have at least one ts mutation in the cdc2 gene and is inhibited in the G2 phase when exposed to 39.5 degrees C. These results suggest that there may be a functional deficiency of pol alpha in senescent human fibroblasts, and this replication factor may be one of the rate-limiting factors involved in loss of the capacity to initiate DNA synthesis in senescent cells.

Decrease in cellular replicative potential in "giant" mice transfected with the bovine growth hormone gene correlates to shortened life span.

Adult mice, (C57BL/6 x Sjl)F1 hybrids, transfected with the bovine growth hormone gene (bGH) grow to twice normal size, but have a mean life span less than 50% that of control siblings without the transgene. The replicative potentials of cells from six different tissue sites (tail skin and ear skin dermal fibroblasts, tail subdermal connective tissue fibroblasts, kidney medulla epithelial cells, bone marrow myofibroblasts, and spleen myofibroblasts) were assayed in vitro using clone size distribution analysis. Cells from all of the above bGH+ tissues produced a smaller fraction of large clones, relative to age-matched controls, in all of these cell types. The loss of replicative potential did not appear to be the result of negative conditioning of the cloning media by the bGH+ cells, and was tightly correlated to the period of accelerated growth in these animals (3-12 weeks), a time when additional GH receptors are expressed.

An age-related reduction in the replicative capacity of two murine hematopoietic stroma cell types.

Two stromal cell types, myofibroblasts and endothelial-like cells, that were identifiable by structural and antigenic specificities, were obtained from murine bone marrow and spleen of young, middle-aged, and old mice of two strains and sexes and grown in liquid culture for 9 or 10 days. As expected, there were more total nucleated cells per organ in the old mice (with larger organs) than in the young mice. However, the concentration of stromal colony forming cells was greater in the young mice, resulting in the number of colony forming cells per organ not being significantly different in most comparisons. The in vitro replicative capacity of the two stromal cell types from both organs in all age groups was determined by clone size distribution assays. In all instances the number of cell doublings achieved was statistically significantly greater in the stromal cell clones from young mice than those from old mice. The cell doubling capacity of the middle-aged mice fell between that of the young and the old mice and in most instances that difference was also statistically significant. It was concluded that these in vitro findings constituted a biomarker of aging in these tissues and that this was significant in relation to previous in vivo and in vitro work by these authors and by others reporting the inferiority of aged bone marrow and spleen stroma to regenerate and to support hematopoiesis.

The cultured diploid fibroblast as a model for the study of cellular aging.

The limited proliferative potential of the cultured human diploid fibroblast is now well established. A number of biological correlates suggest that this culture system is a model for the study of aging at the cellular level. The mechanism(s) that causes the loss of proliferative activity is unknown; the results of some recent studies indicate that specific genes may play a pivotal role in cellular aging in vitro. The extent to which changes in proliferative functions are causally related to aging in vivo is currently under investigation.

The relationship between the rate of entry into S phase, concentration of DNA polymerase alpha, and cell volume in human diploid fibroblast-like monokaryon cells.

We have examined the kinetic relationship between the rate of entry into the S phase in human diploid fibroblast-like (HDFL) monokaryon cells and (1) the concentration of DNA polymerase alpha activity and (2) the cell volume. In the former studies, a first-order dependence between the rate of entry into the S phase and the concentration of DNA polymerase alpha activity was observed, consistent with the enzyme, or a coregulated factor, being rate limiting for this metabolic process. Examination of the nature of the dependence of the rate of entry into the S phase upon cell volume revealed a more complex relationship. The results obtained in studies with synchronized cultures are consistent with the presence of two to three rate-limiting reactants when cell volume is the independent variable. Studies with asynchronous HDFL cell cultures revealed that the smallest cells in the G1 population, presumably the early G1 cells, enter the S phase at an increasing rate as a function of cell volume up to a certain size, beyond which the cells enter at a decreasing rate similar to that observed in the studies with the synchronized cultures. Similar studies examining the relationship between cell volume and the rate of entry into S phase in three established immortal cell lines revealed positive correlation between the rate of entry into S phase and cell volume throughout the size range of the G1 population. This latter observation suggests that the factors involved in the initiation of the S phase may be present in concentrations that are not rate limiting in immortal cell lines.

DNA polymerase alpha and the regulation of entry into S phase in heterokaryons.

We have previously reported that the DNA polymerase alpha activity/unit cellular protein is decreased in late-passage (senescent) human diploid fibroblast-like (HDFL) cultures due to the cellular enlargement associated with in vitro aging. In the studies described here, we have used cell fusion technology to investigate the formal kinetic relationship between the concentration of DNA polymerase alpha and the rate of reinitiation of DNA synthesis in nuclei from senescent cells. Heterokaryons were derived from the fusion of senescent cells to a series of actively dividing cell types with inherently different DNA polymerase alpha activities per cell. A kinetic analysis revealed a first-order relationship between the entry into S phase of senescent nuclei and the concentration of DNA polymerase alpha activity calculated to be in heterokaryons. This result suggests that increases in cell volume may be related to the decline in proliferative activity of late-passage HDFL cells, via "dilution" of factors essential for cellular replication.

Cell enlargement: one possible mechanism underlying cellular senescence.

We previously demonstrated an inverse relationship between the G1 volume of human diploid fibroblast-like (HDFL) cells obtained from foreskin tissue and clonal replicative potential. On the basis of these results, we suggested that one process underlying in vitro senescence is a progressive increase in the mean cell volume of successive progeny within clonal lineages. We now report that the size of HDFL cells, as well as of chick embryo fibroblasts, can be increased in the virtual absence of cell division by culturing at low density and at low serum concentration (0.1-1.0%). Consequent to an increase in cell size, the replicative potential of the cells is reduced to the level of later-passage cells of similar size. By clonal analysis, the populations of enlarged cells contain up to three times as many nondividing cells as do controls. In the enlarged populations, the proportion of cells producing attenuated clones (four or fewer progeny) increases by about 30%, whereas the proportion of cells yielding greater than 32 cells declines by a similar percentage. These observations lead us to propose that replicative potential may be limited by cell size, which in turn may be regulated by a kinetic relationship between cellular growth and cell division cycles.

Proliferative potential of human fibroblasts: an inverse dependence on cell size.

Human foreskin fibroblast-like cells were separated on the basis of DNA content and cell size by fluorescence-activated cell sorting. Subpopulations of "large" or "small" cells with the same (G1) DNA content were clonally expanded and found to contain predominantly nondividing or highly proliferative cells, respectively. From the rate of clonal growth, we deduce that small cells divide faster than large cells. Intermediate-sized cells were found to yield primarily smaller ("attenuated") clones. The clonal data can be incorporated into a previously reported kinetic model of clonal attenuation. This version of the model postulates that small "stem" cells yield larger daughters which have only a limited proliferative potential. We also postulate that a progressive increase in cell size can account for the decreasing concentration of DNA polymerase alpha, which has been reported in older cultures.

Induction of DNA polymerase alpha in senescent cultures of normal and Werner's syndrome cultured skin fibroblasts.

DNA polymerase alpha activity was determined following serum stimulation of early and late passages of human diploid fibroblast-like (HDFL) cultures derived from apparently normal donors (two strains) and from a patient with Werner's syndrome (one strain). Induction of this enzyme was observed in both low passage, actively proliferating cultures and in postmitotic "senescent" cultures from all three strains. The maximal polymerase activity of early and late passage cells of each strain were nearly identical when normalized to the number of cells present. However, the activity of the enzyme was observed to be significantly lower in late passage cultures when normalized to total protein content apparently because of enlargement of the senescent cells. The behavior of Werner derived cells was similar to that of the normal cells. The induction of DNA polymerase alpha in senescent cultures indicates that they retain the capacity to carry out some complex metabolic responses to mitogen stimulation. In addition, these results suggest the possibility that dilution of DNA polymerase alpha and/or other DNA replication factors may play a role in the onset or maintenance of the postmitotic state in the enlarged senescent HDFL cells.

Evidence that a critical threshold of DNA polymerase-alpha activity may be required for the initiation of DNA synthesis in mammalian cell heterokaryons.

The specific activity of DNA polymerase (90% alpha) was determined in nine "neoplastoid" cell lines (Martin and Sprague, 1973) and in three different strains of HDF (human diploid fibroblast-like cells), all examined in logarithmic phases of growth. This was compared to the ability of each cell type to "rescue" (reinitiate DNA synthesis in) senescent HDF cells subsequent to polyethylene glycol-mediated cell fusions. A sharp "threshold" value of DNA polymerase activity was observed below which reinitiation of DNA synthesis in heterokaryons with senescent HDF does not occur. This threshold was especially obvious when the specific activity of DNA polymerase (p moles dTTP incorporated per mg protein or per cell) was divided by the percent of S-phase cells present in each culture as determined by flow microfluorometry. Our results indicate that the specific activity of DNA polymerase-alpha (or some other factor tightly coregulated with it) in "recessive" cell types (those unable to rescue senescent cells) is only about two times this theoretical "threshold" value, and that fusion of recessive cell types to senescent HDF cells reduces the specific activity in the heterokaryon to below this minimum, thus preventing the cells from entering S phase.

Evidence contrary to the protein error hypothesis for in vitro senescence.

A strain of diploid fibroblasts, obtained from the skin of a male infant, was cultured in vitro and cells were tested throughout their lifespan for the appearance of altered glucose-6-phosphate dehydrogenase (G-6-PD) detected either by thermostability studies or by immunotitration. No significant difference was found in the proportion of thermolabile enzyme in 31 young cultures (4.8 +/- 1%, S.E.), in comparison with that in 19 old cultures (4.9 +/- 1%, S.E.). Old cultures had ceased active cell division (49-60 doublings); DNA replication, measured by [3H]thymidine uptake over a period of 24 hours, was limited to less than 5% of these cells. Young cells (5-22 doublings) had a [3H]thymidine labeling index of 75-85%. Titration of G-6-PD activity in extracts of young and old cells with neutralizing antibody directes specifically against G-6-PD failed to detect an increment of enzymatically defective G-6-PD in old cells. The thermostability studies were capable of detecting altered G-6-PD in skin fibroblasts from a female heterozygous for a thermolabile mutant of G-6-PD, and in fibroblasts treated with a proline analogue, azetidine carboxylic acid. The immunotitration technique was also capable of detecting catalytically altered G-6-PD from the thermolabile mutant and G-6-PD inactivated with N-ethylameimide. These findings argue against a protein error catastrophe as the cause of in vitro clonal senescence.