Much is being learned about cell aging and its impact on the actual human aging process--with implications for diseases like cancer. Replicative Senescence By Olivia M. Pereira-Smith and Michael J. Bertram

We are all aware of how our bodies change as we grow older. Like most biological processes, the decline in function of organisms over time is a complex phenomenon. It therefore is not surprising that no two species or individuals within a species age in exactly the same way. The variability in aging between species and between individuals is due to differences in metabolism, environment, and genetic background. This variability makes it difficult to directly relate studies of aging in common laboratory animals like mice and fruit flies to human aging. However, despite this caveat, such animal model systems have provided valuable insights into various aspects of aging at different levels of the complex process. Experiments with humans are not and cannot be as in-depth and controlled, for ethical and moral reasons. Because of this limitation, the information provided by studies of human subjects has only begun to scratch the surface of the underlying mechanisms of human aging. One way around this constraint has been to study the fundamental building block of the human body, the cell, grown in the laboratory by a technique known as cell culture. To make such cultures, human tissue is treated with enzymes that degrade the proteins surrounding the cell, allowing them to be released. They are then placed in specialized dishes containing a growth-promoting solution, the cell culture medium. The methods used to release the cells as well as the culture medium allow selection of the growth of a particular cell type and the establishment of the cell culture. In this way, cells from tissues such as skin, liver, and bone as well as tumors can be studied in rigorous experimental procedures that cannot be performed on humans (Freshney, 1986). Many researchers utilize the cell culture model system, and it has been yielding new insights into the basic molecular processes of human biology and aging. Replicative or cellular senescence was observed and proposed as a model for aging at the cellular level over thirty years ago by Leonard Hayflick (1965; Hayflick and Moorhead, 1961). He studied human fibroblastlike cells obtained from lung and skin. Fibroblast cells are spindle shaped and are important in maintaining the matrix, the material around all cells, in these tissues. Fibroblast cells produce proteins such as collagens and fibronectin, which make the matrix. Hayflick found that when serially cultured, these cells would undergo rounds of divisions, but as the culture aged, the cells were no longer able to divide. In conjunction with the loss of division potential, there were changes in the morphology, the shape and physical appearance, of cells. The cells enlarged significantly, and more space was observed between individual cells. This loss of division potential and the simultaneous change in morphology was termed cellular/replicative senescence. Hayflick did a number of experiments to demonstrate that the cell senescence model was not an artifact of putting cells in culture. These experiments included using culture medium from senescent cells to demonstrate that the medium could still support the growth of young cultures. He also cultured old male cells with young female cells and showed that the culture was eventually composed of only female cells, as the male cells aged first. Although cellular senescence was first described using fibroblastlike cells, all human cell types that can be grown successfully in culture undergo cellular senescence. These types include melanocytes and keratinocytes from skin, adrenal cells from the kidney, T cells from the immune system, and a large number of epithelial-like cells, which come from tissues such as eye, prostate, and breast (Smith and Pereira-Smith, 1996). The in vitro lifespan of different cell types varies, but the maximum lifespan observed has always been fewer than 100 population doublings. It is important to note that this is the number of doublings it takes for all the cells in the culture to become senescent and thus render the culture unable to divide. However, senescent cells are present at all time points during the lifespan of the culture but are a minor part of the cell population when the culture is young. The aging of the culture is not dependent on chronological time, but rather is measured by the number of cell divisions, or population doublings, the culture has undergone. This fact was shown by a number of laboratories that made young cells quiescent, or reversibly nondividing, by removing essential growth factors. The cells stopped dividing at this time, but when they were allowed to resume growth, they went through the same number of doublings as cells that had been growing continuously (Dell'Orco, Mertens, and Kruse, 1974). Also, when cells are frozen in liquid nitrogen after various numbers of population doublings and then thawed and cultured, they undergo close to the same number of doublings as cells maintained unfrozen in culture. Thus, the cells have an internal counting mechanism that allows them to "remember" the number of doublings they have undergone and the number after which they must become senescent. An attractive current candidate for the counting mechanism is telomere shortening, in which dna is lost from the ends of chromosomes at each division (see article by Warner and Hodes, this issue, for details). However, as discussed below, it is clear that there are a number of events--some dependent on telomeres and some not--that result in irreversible growth arrest. The genetic pathways to senescence, and most likely some of the alternate pathways as well, do not appear to be dependent on telomerase (the enzyme that maintains telomere length) and therefore related to telomere shortening. The fact that senescent cells are not dead or dying cannot be overemphasized. Senescent cells are actually resistant to programmed cell death (apoptosis) and in laboratories have been maintained in their nondividing state for as long as three years. These cells are very much alive and metabolically active, but they do not divide. This nondividing state is irreversible by any biological, chemical, or viral agent. At this stage of terminal nondivision, it has been shown that the gene expression cells have undergone global changes compared to those of their younger counterparts. The relationship between the changes in gene expression and cellular senescence has not been definitively established, and it is not known whether any or all of the changes cause senescence or whether senescence results in the changes in gene expression. Gene expression changes that could potentially induce senescence include a repression of cell-growth- inducing transcription factors (Dimri and Campisi, 1994). However, along with this repression of growth inducers is an activation of the cell cycle inhibitors, p21 and p16, which are more likely the genes that act to induce cell senescence and in fact are the end products of genetic programs that lead cells to senescence (Smith and Pereira-Smith, 1996). Pathways to Senescence A number of pathways to senescence exist. Genetically defined pathways. Immortal cells have escaped the cell senescence pathways and can grow indefinitely in culture. These cells can be derived from tumors (although not all tumors have totally escaped senescence), transformed with viruses, or induced with chemical agents. To elucidate the genetic pathways that lead to senescence, research has focused on the differences between immortal cells and normal cells. Early studies showed that fusion of immortal cells with normal cells resulted in a senescent phenotype, establishing that senescence, or the normal phenotype, is dominant over cellular immortality (Pereira-Smith and Smith, 1983). This finding demonstrates that the immortal phenotype is due to recessive genetic mutations in the pathway(s) that lead to senescence. To extend this idea, an immortal cell complementation study was undertaken whereby cells from different immortal cell lines were fused with each other. If hybrids from fusion of two different immortal cell lines had an immortal phenotype, this would indicate that they must have a lesion in the same senescence gene or gene pathway. If the hybrid phenotype of the fusion was regaining normal senescence, the immortal parental call lines must have mutations in different senescence genes or gene pathways. In this way, more than forty immortal human cell lines have been tested, and all fall into one of four complementation groups, A, B, C, or D. These results indicate that there are four genes or gene pathways that lead to senescence; inactivation of any one of them is a necessary step toward immortality (Pereira-Smith and Smith, 1988). Identification of the genes underlying the complementation groups has been a slow, arduous process and, at this point, one with limited success. To date, the chromosomes that harbor the senescence genes for three of the complementation groups have been identified using the technique of microcell fusion, by which single, individual human chromosomes have been introduced into cells lines representing the different complementation groups and the resulting phenotype ascertained. Chromosomes 1, 4, and 7 have been identified as carrying cell-senescence-related genes (Ning et al., 1994; Henslek et al., 1994; Ogata et al., 1993). morf4, the senescence gene on chromosome 4, was recently cloned (Bertram et al., 1999). Alternate pathways. Recently, there have been a number of studies in which cells have been treated with various agents and induced to quit the cell cycle and enter a senescentlike growth arrest (Young and Smith, 2000). In these studies, some changes known to occur when cells enter senescence after undergoing replication have been studied and found to occur. The most consistent is the enlarged morphology reminiscent of a senescent cell. It is thought that the mechanism of action in these cases is through production of changes that affect the cell and also global gene expression. However, more complete characterization of these growth-arrested cells is needed to confirm how similar to senescent cells they truly are. Nonetheless, it seems reasonable to assume that, depending on the signals a cell receives, it can choose to enter senescent growth arrest and remain viable or die. The cells enter an arrested state very rapidly, less than a week after treatment, suggesting that the time-consuming process of telomere shortening is not always involved. Correlation of in vitro Cell Senescence to in vivo Aging The big question is whether cellular senescence does play a role in human aging and, if so, in what capacity. There are three areas in which cellular senescence could affect processes in a living organism (in vivo processes). The first is loss of division potential in tissues like bone, skin, liver, and gastrointestinal epithelium. In these examples, the questions are whether, with time, we are running out of cell divisions or whether the time increased for cell division to occur is compromising these organisms, leading to loss of function. An example of in vivo senescence could be the decline in immune function that is observed with aging (Miller, 1996). There is evidence that this decrease in immune function is due at least in part to loss of division response of T cells to antigens such as viruses. Similarly, one of the common problems of aging, osteoporosis, could well be due to inadequate division of osteoblasts, the cells that lay down bone (D'Ippolito et al., 1999). This phenomenon would allow excessive bone resorption by the osteoblast, resulting in lower bone density. In fact, in culture, osteoblasts from older people grow less well than those from young individuals. But in the case of other tissues like skin, the number of population doublings that cells can achieve in culture would provide enough cells for as much as 200 years, which is beyond our known lifespan. Another attractive role for cellular senescence in vivo is as a cancer control mechanism. Since the cells can only go through a certain number of divisions as a primary tumor grows, they will enter senescence unless they escape growth control by one of the four pathways described above. This state of senescence will keep tumors from growing and thus prevent the spread of the cancer. There is also evidence that senescence is sometimes induced in cells that have been damaged, as an alternative to cell death and damage. In this way, cell senescence is often in response to the induction of tumors and would thus act as a line of defense against cancer progression. Senescent cells may themselves be a disruptive and destructive force in some instances. Senescent fibroblast cells have been shown to accumulate with age in skin biopsies (Dimri et al., 1995). For fibroblasts, in particular, senescence is accompanied by increased expression of degrading enzymes like collagenase and stromelysin. These enzymes potentially could be detrimental to the environment outside and around the cell and thus could affect all the other, nonsenescent cells nearby. In skin, for example, this could lead to a loss of integrity, resulting in the wrinkles and loss of elasticity common with aging. Another possibility is that this change in tissue composition would make tumor metastasis easier and thus might explain why tumor formation increases with age. However, these possibilities are based on more conjecture than fact, and it remains to be proven that replicative senescence does affect in vivo aging. Better markers of senescent cells are needed to allow for an unequivocal demonstration that this is the case. The problem has been that many of the genes that define senescence are also expressed in nondividing, quiescent cells. Perhaps the complementation group-related genes would act as better markers of the true senescent state. Summary The study of cellular/replicative senescence has clearly advanced the understanding of basic biology, cell cycle control, and transcriptional regulation of genes. The fact that telomerase can extend the number of population doublings that cells can undergo provides a powerful tool for ex vivo "gene therapy," in which a person's own cells can be genetically manipulated to compensate for a defect and then returned to the individual (see Hornsby, this issue, for details). There is clearly a lot left to be done to demonstrate the role of this phenomenon during in vivo aging, but as critical genes are identified, they can be analyzed in animal models and human aging models to better define the impact of cell aging on the actual human aging process. Olivia M. Pereira-Smith, Ph.D., is professor, Division of Molecular Biology and Departments of Medicine and Cell Biology, Baylor College of Medicine, HuYngton Center on Aging, Houston, Tex. Michael J. Bertram, Ph.D., is assistant professor, Department of Medicine, Division of Endocrinology and Metabolism, University of Alabama, Birmingham. References Bertram, M. J., et al. 1999. "Identification of a Gene That Reverses the Immortal Phenotype of a Subset of Cells." Molecular and Cell Biology 19: 1479­85. Dell'Orco, R. T., Mertens, J. G., and Kruse, J. F. J. 1974. "Doubling Potential Calendar Time, and Donor Age of Human Diploid Cells Inculture." Experimental Cell Research 84: 363­6. Dimri, G. P., and Campisi, J. 1994. "Altered Profile of Transcription Factor-Binding Activities in Senescent Human Fibroblasts." Experimental Cell Research 212: 132­40. Dimri, G. P., et al. 1995. "A Biomarker That Identifies Senescent Human Cells in Culture and in Aging Skin in vivo." Proceedings of the National Academy of Sciences USA 92: 9363­67. D'Ippolito, G., et al. 1999. "Age-Related Osteogenic Potential of Mesenchymal Stromal Stem Cells from Human Vertebral Bone Marrow." Journal of Bone and Mineral Research 14: 1115­22 Freshney, R. I. 1986. "Animal Cell Culture: A Practical Approach." Practical Approach Series 1: 1­226. Hayflick, L. 1965. "The Limited in Vitro Lifetime of Human Diploid Cell Strains." Experimental Cellular Research 37: 614­36. Hayflick, L., and Moorhead, P. S. 1961. "The Serial Cultivation of Human Diploid Cell Strains." Experimental Cellular Research 25: 585­621. Hensler, R. J., et al. 1994. "A Gene Involved in Control of Human Cellular Senescence on Human Chromosome 1Q." Molecular Cellular Biology 14: 2291­7. Miller, R. A. 1996. "The Aging Immune System: Primer and Prospectus." Science 273: 70­4. Ning, Y. H., et al. 1994. "Cloning of Senescent Cell-Derived Inhibitors of dna Synthesis Using an Expression Screen." Experimental Cellular Research 211: 90­8. Ogata, T., et al. 1993. "Chromosome 7 Suppresses Indefinite Division of Nontumorigenic Immortalized Human Fibroblast Cell Lines kmst-6 and susm-1." Molecular and Cellular Biology 13: 6036­43. Ogata, T., et al. 1993. "Genetic Complementation of the Immortal Phenotype in Group D Cell Lines by Introduction of Chromosome 7." Japanese Journal of Cancer Research 86: 35. Pereira-Smith, O. M., and Smith, J. R. 1983. "Evidence for the Recessive Nature of Cellular Immortality." Science 221: 964­6. Pereira-Smith, O. M., and Smith, J. R. 1988. "Genetic Analysis of Indefinite Division in Human Cells: Identification of Four Complementation Groups." Proceedings of the National Academy of Sciences USA 85: 6042­6. Smith, J. R., and Pereira-Smith, O. M. 1996. "Replicative Senescence: Implications for in Vivo Aging and Tumor Suppression." Science 273: 63­7. Young, J., and Smith, J. R. 2000. "Epigenetic Aspects of Cellular Senescence." Experimental Gerontology.

Generations Table of Contents

American Society on Aging
833 Market St., Suite 511
San Francisco, CA 94103
www.asaging.org
info@asaging.org