Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Mode] Systems for the Evaluation of Toxic Agents Affecting Aging or Age-Related Diseases . The development of mode! systems in the field of aging is at about the same stage today as the development of mode! systems to detect potentially mutagenic environmental chemicals was in 1970. Screening tests had shown that potentially mutagenic chemicals belonged to a wide variety of chemical classes, for example, foods, drugs, cosmetics, agricultural chemicals, and household chern~cals. In the first phase of the development of short-term screen- ing tests, appropriate mode! systems were used to obtain infor- mation on the various mechanisms that produce genetic lesions and to identify the types of genotoxic agents that would produce gene mutations, chromosomal aberrations, and aneuploidy. Those models were almost exclusively in vitro assays that used organisms ranging from bacteria to mammalian cells in culture. The second phase was the identification of mammalian models that could be used to evaluate chemicals that had positive results in in vitro tests to determine the effects of these chemicals on somatic celb and germ cells. The later tests were selected to provide a comprehensive screen in the mammal that could generate a data base suitable for predicting human response qualitatively and quantitatively. The development of short-term tests for mutagenicity provides 145
146 AGING IN TODAY'S ENVIRONMENT a useful example for the development of appropriate model shy terns to identify environmental chemicals that affect aging. The procedure in using mode! systems to evaluate the effects of envi- ronmental chemicals on aging is first, to identify short-term tests that can be used to screen environmental chern~cals for their po- tential to affect aging processes or age-associated diseases; second, to screen chemicals; and third, to test those chemicals that had positive results in short-term tests in mammalian model systems to develop a data base for predicting human responses. CONSIDERATIONS IN CHOOSING AND DESIGNING MODEL SYSTEMS The general lack of information about the fundamental nature of aging processes poses problems for risk assessment of potential environmental agents that promote aging. In the absence of reli- able information, a prudent course might be to use multiple model systems. The considerations in choosing model systems for as sessing relationships between environmental agents and aging or age-related diseases include: Length of life or of assay. . Previous use of the model in the study of aging and toxi- cology. . Knowledge of and ability to control for adventitious m croorganisms. netic analysis. . Knowledge of genetic characteristics and suitability for ge- Capacity to maintain defined environmental conditions. Knowledge of pathologic changes associated with aging. Ready availability of the cells or organisms. Economic feasibility. Widespread use in other biologic disciplines. Relevance to aspects of human aging. Life span is by far the most widely used marker in assessing the effectiveness of experimental interventions into the aging pros cesses. If the end point is longevity, life-table data are required, including life expectancy at birth and life span of the animal model to be used. For many strains of mice and rats and for many inver- tebrate species, sound life-table data are available. Species with short life spans are distinctly advantageous for such uses. A short
MODEL SYSTEMS FOR EVALUATING TOXIC AGENTS 147 life span enables the investigator to study the animals longitudi- nally throughout their lives under well-defined conditions and to analyze the data with the kiln of undertaking further studies within the time frame of the investigator's scientific "life span. It must be emphasized, however, that age-associated changes might occur in a long-lived species like humans that do not occur in short-lived species. Obviously, when such changes are to be studied in an animal model, the time required for the changes to develop must be considered in relation to the an~mal's life span. Many invertebrates have life spans of only a few days. Such organisms might be very useful in the early stages of a comprehen- sive toxicity-testing regunen. If the assay involves measures other than life span, it should also be of short duration. Fast assays cost less than slow ones, yield results rapidly, and make it possible to assay many compounds in a small space and a short tone. The importance of keeping a mode} free of infectious diseases is well illustrated by the reports of Paget and Lemon (1965), who compared the longevity characteristics of conventionally main- ta~ned Wistar rats with those of Wistar rats kept specific-pathogen- free in a barrier facility. Both the median length of life and the life span of the specific-pathogen-free rats were markedly greater. The data suggest that infectious disease might distort the assessment of the aging processes and thus compromise the value of the model. The genetic characteristics of the animal mode] should be known and the genotype should be stable to ensure reproducibility and to ease interpretation in aging studies. An example of the problems that can be encountered otherwise is illustrated by the use of the popular ~outbred" strains of rats (e.g., the Sprague- Dawley and Wistar strains) for aging research. These strains are maintained by randomly (or not so randomly) mating members of the stock; a procedure likely to result in each supplier having a stock with genetic characteristics different from those of other suppliers. The genetic heterogeneity and the investigator's lack of awareness of its existence might result In erroneous interpretations, for instance, if data are obtained on young rats from one supplier and old rats from a different supplier. Many invertebrate species have been used extensively in so- phisticated genetic analyses involving both classical and molecu- la~ genetics. For example, the metazoan organism with the best understood genetics is Drosophila melanogaster. The extensive differences in life expectancy and life span reported in different
148 AGING IN TODAY'S ~VIRO~NT stuclies on D. melanogaster relight wise In part from genetic drift during laboratory propagation. Genetic drift can be avoided by frequent referral to a common source or by maintenance of stocks in a nongrowing state. The availability of a common reference strain that can survive indefinitely while frozen, which eliminates genetic drift, and its sophisticated genetics make Caenorha6ditis elegans a system of choice also. Riley (1981b) demonstrated that it ~ important to define and control all environmental conditions. He found that over 65%0 of conventionally housed female C3H/HeJ mice had mammary tumors at the age of 400 days, whereas fewer than logo of the mice protected from the noise, odors, and other stressors common to conventional anneal facilities had these tumors at the same age. Riley related that finding to the difference in plasma corticosterone concentrations, which were 15~500 ng/m! in the conventionally housed mice and ~35 ng/m] in the mice housed in the protected environment. A major advantage of invertebrate organisms is their ability to grow under completely defined environmental conditions. More- over, the fungi and C. elegans can even be grown in chemically defined media, albeit with some alteration in maximal growth ki- netics and other life-history traits. This capability can simplify nutritional studies on aging. The use of cell, tissue, and organoid cultures as mode} systems for the investigation of the aging effects of chemical, physical, and nutritional agents on somatic cells has a number of attractive features. Tissue-culture studies can simplify interpretation of the results of in viva studies by avoiding such variables as the effects of the neural, endocrine, and ~rnmune systems and the nutritional, microbial, and pathologic status of the host animal. Knowledge of the pathologic changes that occur with age in an animal model is essential for both the design and the interpretation of aging studies. Published data on age-related pathology are available for some, but not all, animal models. In addition, the pathologic causes of senescent death in most invertebrates have not been well explored. The immediate availability of a source of aged animals, main- tained in carefully specified environments, would simplify many studies. The lack of availability of aged animals of short-lived
MODEL SYSTEMS FOR EVALUATING TOXIC AGENTS 149 strains does not present a serious problem, because in many re- gards it is preferable for the investigator to be in control of the animal's environment over its whole life. Such considerations are usually of relatively minor importance in invertebrates. Many species are readily available from a variety of sources, including stock centers and individual laboratories (see Appendix A). Many of these organisms can be shipped by regular mail without special precautions, and several species, including fungi and C. elegans, can be storer} indefinitely in the laboratory without passage. An advantage of the invertebrates is their relatively low cost. For example, a complete life table can be constructed from a population of 100 nematodes for a cost of about $50, including overhead, technician time, and data analysis. A comparable study using mice, however, costs at least $10,000, and about twice that for using rats. Such cost considerations are a major force in the development of biologic markers of aging other than life span. All strains and species mentioned here are widely used in other biologic disciplines. In particular, Neurospora crassa, C. elegans, and D. metanogaster are among the most widely studied metazoan organisms. The use of widely studied organisms is an important consideration that was covered in detail in a previous report of the National Research Council (1985~. For a toxicity-test system to be truly valuable, it must closely mimic the toxic response typical of an aging human population or must represent a relevant end point of human aging or an age-related disease. At our current poor level of unclerstanding of aging, it is impossible to make completely informed decisions about the validity of mode} systems. Many fundamental physio- logic and molecular processes in humans are also present in many invertebrates ~d other vertebrates, and there is no a priori reason to expect aging processes to be different. Each model system must be approached on its own merits, and validity assessment must be based on background information on the disease state, the mode! system, and the particular assay under analysis. EXAMPLES OF MODEL SYSTEMS ~ Vitro Modem The ability to maintain and grow cells and tissues outside the
150 AGING IN TODAY'S ENVIRONMENT body has progressed to the point where the effects of environmen- tal agents on living cells can be studied in vitro. For the purposes of screening environmental agents, the use of such assays can result in substantial economnes and efficiencies particularly important when alternative studies require aged animate of several long-lived species or large numbers of dos~response and drug-interaction ex- periments. Stock ceils and, in some cases, tissue explants can be cryobiologically stored in large amounts. This permits repeated assays with comparable materials and the sharing of common stock materials by numerous laboratories. Moreover, such stocks can be used to investigate cell-cell interactions, such as metabolic cooperation and metabolic transformation. Fmally, tissue-culture approaches can substantially reduce the numbers of animals re- quired for experimentation. Such methods cannot, however, be expected to elirn~nate the need for experimentation with intact animals. There are three general categories of methods: organoid cul- tures, tissue explants, and cell culture. Organoid cultures involve the short-term maintenance of viable intact segments of tissue, for example, the full thickness of a segment of aorta. ~ tissue explants, the early migration and proliferation of epithelioid and fibroblastoid cell types can be observed. Cell cultures are of three general types: ~ Primary clones and cultures, that is, proliferating colonies and mass cultures of cells taken directly from the animal, usually after enzymatic dispersion of biopsied tissue. . Established, serially passaged cultures with relatively re- producible cycles of growth in early phases, but with limited replicative life span, and with a genetic makeup reflecting that of the donor animal. "Transformers cell cultures with indefinite replicative pm tential and generally with altered genetic makeup. . Each of the materials described could prove useful for stud- ies of the effects of environmental agents on aging. One general approach would be to explore the toxic effect of an agent as a function of donor age so as to detect unusual susceptibilities of the cells and tissue of aged subjects. Another general approach would be to use the tissue-culture methods after in viva treatments. If a set of behaviors or phenotypes were observed with tissue from
MODEL SYSTEMS FOR EVALUATING TOXIC AGENTS 151 young, treated subjects that proved to be comparable to that oW served with untreated tissue from old anunals, an effect of the in viva treatments on aging processes could be inferred. An entirely different experimental paradigm could be based on the hypothesis that established cultures with finite replicative life span recapitulate the natural history of comparable cell types in viva in aging animals the well-known sin vitro mode! of cellu- lar agings first developed by Hayflick and Moorhead (1961~. The appropriateness of such models for the study of aging is contro- versial, but considerable evidence supports the proposition that t.h~ z`~.~.~nll;~tion Of growth oh~erved in vitro corresponds to events _ O · · . OccurRlilg In VlVO. Of special interest would be the evaluation of agents that exhibit unusual toxicity in putative stem cells or that accelerate the terminal differentiation of such stem ceils, because an excessive depletion of stem cells might seriously compromise the regenerative potential of tissues in aging subjects. In such studies, it wall be important to investigate a variety of cell types. Most research in this field has concentrated on the in vitro aging of cultures of fibroblastoid cells established from the fetal lung or from the dermis of individual subjects of various ages. The precise cell types of origin of such cultures are not clear. Thus, it will be difficult to compare age-related changes in viva with those observed in vitro. Considerable progress has been made in the culture of other cell types, including epidermal keratinocytes, epidermal melano- cytes, lenticular epitheliai cells, renal epithelial cells, chondrocytes, arterial and venous endothelial cells, vascular smooth muscle cells, skeletal muscle satellite cells and myoblasts, erythroid stem cells, myeloid stem cells, T arid B lymphocytes, glial cells, and retinal epithelial cells. Much of this progress has been associated with the improved characterizations of optimal growth media, especially of subsets of growth factors. In some instances, growth in chern~cally defined or nearly defined media has been achieved. A final experimental paradigm would be to use, in culture, postreplicative terminally differentiated celb to investigate agents for their potential to accelerate age-related alterations observed in viva. Such cell types might be derived through in vitro terminal differentiation, for example, or normal or transformed embryonic neuroblasts or myoblasts. A major concern, however, would be the extent to which the experimental milieu reflected in viva con- ditions.
152 AGING IN TODAY'S ENVIRONMENT Nonmam~an Animal Modem Numerous invertebrate species fulfill ~nany of the criteria out- lined earlier and have been widely used as modem for the study of aging. Some have also been used In studies of genotoxicity. More important, several have already been used to examine toxic effects on aging or age-related processes. The most widely used and perhaps the only commonly ac- cepted biomarker of aging is life span (Comfort, 19793. Life span has also been the end point most often assayed in invertebrates, although other end points, such as accumulation of lipofusc~n and the end of reproductive ability, have also been used. Because of their short life spans, relative ease of use, and relatively low cost, invertebrate organisms will likely be important ~ the initial phases of a test system to detect environmental toxins that might affect aging or age-related diseases. Life-table data are a sine qua non of aging research and are therefore essential in a test system to assess toxic effects on aging or age-relatec! diseases. Because invertebrates are often relatively short-lived, it is easy to collect reliable data on the invertebrate of choice. Because such data might not be representative of other genotypes or environments, however, it is often best to choose an organism on which adequate life-table data have already been collected by many investigators over an extended period and under a variety of experimental conditions. Invertebrates for which adequate life tables are available in- clude fungi, especially N. crassa and Podospora anserine (Esser and Bockelman, 1985; Munkres, 1985~; protozoa, such as Parame- cium telraurelia and Tetrabymena pyriformis (Sm~th-Sonneborn, 1985a,b,c); rotifers (Barrows and Kokkonen, 1985~; the nematodes C. elegans, Turbatrix aceti, and C. briggsce (Johnson, 1984; John- son and Simpson, 1985; Russell and Jacobson, 1985~; and insects (tints, 1985a), especially D. melanogaster (Baker et al., 1985; Lints, 1985b), but also Musca domestics (Cheeky, 1985; Sohal and McArthur, 1985), Ha6ro6racon juglandis, and Tribolium confusum (Soliman, 1985~. Several other invertebrate animal species, less widely studied (tints, 1985c; Mitchell and Johnson, 1984), and plants (Nooden and Thompson, 1985), whose mode of senescent action has been widely explicated, have also been used as models for the study of aging. Of the organisms that have been widely used in aging research,
MODEL SYSTEMS FOR EKALUAT~G TOXIC AGENTS 153 four are of particular interest because they fulfill many of the criteria outlined earlier: P. tetraureiia, which has a replicative life span of about 200 cell divisions, encompassing about 40 days (Smith-Sonneborn, 1985b,c); C. elegans, which has a mean life span of 1~30 days, depending on temperature and food (Johnson and Simpson, 1985~; D. melanogaster, which survives an average of about 40 days (Baker et al., 1985~; and M. domestica, which has a life span of only about 2 weeks (Cheeky, 1985~. Only two (the nematode C. elegans and the fruit fly D. metanogaster) are widely used in other kinds of biologic research and are amenable to the wide range of molecular and genetic analyses currently available in the biotechnolog~c armory. Some protozoa have little ability to continue mitotic replica- tion in the absence of mating and so have been used to study the phenomenon of finite proliferative life span, usually termed clonal aging (Sm~th-Sonneborn, 1984~. Protozoa display a spec- trum of finite replicative divisions among different species, ranging from about 40 divisions to apparent clonal immortality. Probably the best-studied species in aging research is Paramecium aure- lia, whose age-related morphologic changes have been described during clonal aging, including rn~cronuclear, macronuclear, and cy- toplasmic changes. Functional changes in the rate of macromolec- ular syntheses have been reported. Paramecium has been used in assessing the effects of environmental insults, particularly radia- tion, on clonal life span and replication rate (Sm~th-Sonneborn, 1985c). Classical genetics is available for many protozoa, includ- ing Paramecium. Molecular genetic analysis is well developed in some species, but is almost completely lacking in Paramecium, and studies are limited to a few laboratories. Nematodes, particularly C. elegans and T. aceti, have been widely studied as models of metazoan aging (Johnson, 1984; John- son and Simpson, 1985; Russell and Jacobson, 1985~. The somatic cells of adult nematodes are all postreplicative; mean life spans are a few weeks; and kinetics of death display the Gompertzian increase in mortality seen in higher metazoans. A variety of mor- phologic, behavioral, physiologic, and molecular changes occur over the life span of the nematode (Johnson and Simpson, 1985~; some uremic changes observed in the mammalian aging processes, such as loss of general motor ability (Bolanowski et al., 1981; Johnson, 1987) and lipofuscin accumulation (Clokey and JacoW son, 1986~. The animals can be grown on a simple bacterial diet
154 AGING IN TODAY'S ENVIRONMENT or in a completely defined medium (Johnson, 1984; Russell and Jacobson, 1985~. C. elegans is the object of study in about 50 laboratories throughout the world. Sophisticated classical and molecular ge- netic analyses on this organism are available (Russell and JacoW son, 1985), including genetic transformation (Fire, 1986~. The entire cell lineage, from one-cell stage to adult, has been described (SuIston et al., 1983~. Mutagenesis by transposable elements and DNA transformation are available. A range of genetic variants with lengthened life span are also available (Friedman and John- son, in press; Johnson, 1987~. Nematodes have been widely used as models in genetic tox- icology, as well as in the ascertainment of the effects of drug treatments on life span (Johnson, 1984; Johnson and Foltz, 1987~. It would be inappropriate to expect to mode! all aging processes of humans in any invertebrate. For example, all somatic cells in C. elegans adults are postreplicative and therefore do not mimic replicating mammalian cells. D. melanogaster, the most widely studied invertebrate species, has been the most widely used in aging research (Baker et al., 1985~. Drosophila has life spans of a few months and can be main- ta~ned in the laboratory conveniently under well-defined growth conditions. A wide variety of morphologic, behavioral, physiologic, and molecular changes with age have been described (Baker et al., 1985), some of which parallel changes observed in mammals. Most important, the fruit fly has been intensively studied for over 70 years as a genetic mode} and for over 60 years as a general genetic mode! of metazoan aging (tints, 1978~. Studies have been completed with a wide variety of dietary supplements, environmental insults, and potential toxins to assess the effect of these substances on life span (Baker et al., 1985~. Sophisti- cated molecular genetic techniques are available, including almost routine transposable-element mutagenesis and molecular cloning (Spradling and Rubin, 1982~. More is known about the genome of this organism than about that of any other higher metazoan, including the human. Moreover, Tong-lived strains of Drosophila have been derived by selective breeding (I,uckinbill et al., 1984; Rose, 19843. Except for Drosophila, the house fly Musca fdomestica is the most widely used insect mode} for aging research (tints, 1984~. Musca is easy to maintain, and mean life spans are around 20
MODEL SYSTEMS FOR EVALUATING TOXIC AGENTS 155 days. Molecular assays can be performed on single flies. A variety of morphologic, behavioral, molecular, and physiologic charac- teristics display changes over the life span (Cheeky, 1985~. The free-radical theory of aging has been most directly examined in Musca with detailed physiologic and molecular assays (Cheeky, 1985; Sohal, 1981~. Drugs have been used to modify the life span and physiology in an effort to test the free-radical mode] further. Unfortunately, almost no classical or molecular genetic analysis of the house fly is available (Cheeky, 1985~. The fungi N. crassa and P. anserine have been widely ex- ploited both in aging research and in classical and molecular genetic analyses. The haploid nature of the fungi makes them especially amenable to the identification of some types of muta- tions, and both of these fungi have been used to characterize the only senescence phenomenon understood at the molecular level (Esser and Bockelman, 19853. Most molecular genetic techniques are also ava~iable for use with these fungi. Many invertebrate species can be grown under rigorously de- fined culture conditions. C. etegans con be cultured on chemically defined media and cells of D. me~anogaster can be grown in cul- ture. Although many invertebrates have been examined in aging research, D. meianogaster and C. elegant are widely used not only in aging but also in other fields, and thus should be emphasized. The lack of adequate pathology is a major limitation in the use of many invertebrates. For example, only sparse pathologic descriptions at death are available for Drosophila at the elec- tron microscopic level (Baker et al., 1985) and for M. domestics (Cheeky, 1985; Sohal and McArthur, 1985~. Only a minimal de- scription, at the electron microscopic level, of senescent changes in C. elegans has been made (Johnson, 1984~. Invertebrates are phylogenetically far removed from mammals and might therefore slider from them in fundamental ways. Al- though many studies of basic biochemical and physiologic events have shown that these processes are conserved over large phylo- genetic distances, it is clear that not all physiologic processes are conserved. The lack of information on the nature of basic aging processes in any metazoan makes it impossible to determine, a priori, the relevance of any organismic mode! to human aging. Nevertheless, in some important applications, such as assessment of toxic effects on life span and modification of life span by toxic
156 AGING IN TODAY'S ENVIRONMENT agents, invertebrates are useful modem and are cost-effective al- ternatives to mammalian modem. Mammalian Modem Most mammals do not fully meet the criteria of a short life span, relative ease of maintenance in a defined environment, wide use In biologic research, and suitability for a wide variety of molec- ular and genetic analyses. Mice and rats best fulfill these and other criteria for aging research (National Research Council, 1981b), and indeed they have been and are widely used In aging research. Many vertebrate species, such as rabbits (National Research Council, 1981b), lack good lif~table data, although they are often used as models for atherogenesm research. It is difficult to obtain a rabbit more than 5 years old, not because of spontaneous death, but because breeders usually kill the animate by that age. Two studies have addressed the longevity characteristics of rabbits; one indicated a life span of 8 years (Weisbroth, 1974) and the other of 13 years (Flower, 1931~. Before rabbits can be electively used in aging research, further information concerning their longevity is necessary. Even when excellent life-table data are available, in,restig~ tors often fad! to make use of them ~ designing their aging stud- ies. Prominent biochemists and metabolic physiologists cornrnonly draw conclusions about age-related changes in a biochemical pros cess on the basis of a study limited to 2-month-old and ~month-old rats in a strain with a life span of 48 months. Although such stud- ies are of value, a broader range of ages should be studied if the influence of age on biochemical activity is to be aclequately defined. The value of published pathologic data is iilllstrated by the following example. The male Fischer 344 rat ~ a popular mode! for aging studies. More than 50~o of these rats have testicular interstitial cell tumors at the age of 18 months, and by the age of 24 months almost all have the tumors (Coleman et al., 1977; Yu et al., 1982~. That information is critically important in designing studies on the aging of the male reproductive system. It is also important to carry out pathologic analyses in the same animals on which physiologic or biochemical measurements are being made. For instance, if the concentration of sequin parathyroid hormone increases in some but not all old male Fischer 344 rats, it Is important to know whether this increase could be secondary to a
MODEL SYSTEMS FOR EVALUATING TOXIC AGENTS 157 coexisting disease process, for example, a severe grade of chronic nephropathy. Rats have the advantage of providing more biologic material per animal than mice without markedly increasing the ma~nte- nance cost or space requirement. A disadvantage is the small number of inbred strains available (National Research Council, 1981b). The Fischer 344 strain is the major inbred rat strain that has been used in aging research. It has several advantages in addition to genetic homogeneity: it does not become obese with advancing age (Bertrand et al., 1980a), extensive data on its age-related disease processes have been published (Coleman et al., 1977; Maeda et al., 1985), and there is a sizable body of life-table data (Masoro, 1980; Yu et al., 1982, 1985~. The major disadvantage is the occurrence of progressive chronic nephropathy and renal failure in animals fed ad libitum at a relatively early age (Maeda et al., 1985~. Restricting the rat strain to 60% of the ad libitum food intake prevents the pros gressive chronic nephropathy. The National Institute on Aging is developing Brown-Norway, Fischer 344 Fat hybrids for the purpose of partially circumventing chronic nephropathy. It is already clear that a controlled and defined dietary progam is necessary in aging studies. Other rats that have been extensively used for aging research are the Sprague-Dawley, Waster, and Long-Evans (Masoro, 1980~. They are not inbred strains and are not genetically homogeneous. It is more appropriate to call them stocks than strains (National Research Council, 1981b). Life-table data are available (Masoro, 1980~. A comprehensive report (Anver et al., 1982) on the age- related disease processes of the male Sprague-Dawley rat is avail- able. A disacivantage of the male Sprague-Dawley and Wistar rats, compared with male Fischer 344 rats, is that they become obese with advancing age (National Research Council, 1981b). Because they are small, mice are relatively inexpensive to maintain and clo not require a large space. Moreover, many inbred strains are available, and some are even available as aged animals (see Appendix A), which facilitates genetic exploration of aging (National Research Council, 1981b). A disadvantage of mice is that only a small amount of biologic material can be obtained from a single animal. Good life-table data are available on sev- eral strains (National Research Council, l9Blb): C57BI,/6N Nia males, BAI`B/cN Nia males, CB6F males, C57BL/6] females,
158 AGING IN TODAY'S ENVIRONA~3NT DBA2/] females, B6D2F females, C57BL/63 males, DBA2/] males, B6D2F~ males, C3Hf/B~ females, BALB/c An B3f fe- males, C3CF females, C3Hf/B] mates, BALB/c An B6f males, and C3CF~ males. Inbred strains often suffer from a single major dmease process (e.g., cancer of the liver or chronic nephropathy), and the presence of this major disease process in most if not all the animals compli- cates gerontologic interpretation. Fit hybrid strains offer a partial solution to the problem and can be readily produced in large num- bers. McClearn et al. (1970) have developed a more genetically heterogeneous stock of mice that is systematically maintained. Their procedure yields a genetically stable population without the drawback of inbred strains and thus probably provides a mode! for aging research in which genetic variation is readily accessible. Recombinant inbred strains (Bailey, 1981) should also be more widely used. Because of the genetic identity of each member of a population, some individuals can be sacrificed while identical siblings are maintained for survival and analyses on living animal (Johnson, 1987~. If old animals are available from suppliers, the investigator must also make certain of the availability of accurate "formation on the lifetime environment of the animals, including dietary his- tory and data on the monitoring of infectious disease. Usually, such data are not available on long-lived animal models of ad- vanced age. Indeed, the lack of well-defined animals and the cost of their purchase and maintenance have made it difficult to carry out aging research on adequately characterized long-lived animal models. Reasonable life-table data are available on the beagle and the thoroughbred horse. With regard to nonhuman primates, the two principal mod- els that appear to be emerging are the pig-tailed macaque mon- key (Macaca nemestrina) (Bowden, 1979) and the rhesus monkey (`Macaca mulatta) (Davis and Leathers, 1985~. It might be de- sirable to develop aging cohorts of the chimpanzee because, from a biochemical and genetic point of view, chimpanzees are closely related to man; but practical considerations limit the feasibility of this model. The availability of tissues and fluids from a selection of mam- malian species of contrasting life spans, ranging from the short- lived murine species to the long-lived primates, would permit a
MODEL SYSTEMS FOR EVALUATING TOXIC AGENTS 159 systematic comparative approach to the study of aging. For ex- ample, it would permit the differentiation of markers that appear to be related primarily to chronologic rather than biologic age, and thus provide a stronger rationale for the choice of assays for the effects of agents that might alter aging processes. Epidemiologic Models The expression "experiments of nature" refers to contrasts in the exposure of human populations to toxic agents that provide an opportunity to assess the impact of toxic exposures on the risk of disease and death in the human population. Many reports of acute and chronic releases of toxic agents into the human environ- ment gave rise to marked contrasts In toxic exposures and disease outcomes. They include the mercury contamination of Minimata Bay in Japan, the atomic blasts in Hiroshima and Nagasaki, and smog episodes in London. The first of the weli-publicized air-pollution incidents that re- sulted in marked increases in illness and death, mostly among the elderly, occurred in the Meuse River Valley in Belgium in 1930. This heavily industrialized area was seriously affected when ac- cumulating air contaminants, trapped by an inversion resulting from a thick, cold fog, caused 60 deaths and illness in thousands of residents. The incident at Donora, Pennsylvania, in 1948 resulted from a similar inversion that covered a wide area of the northeast- ern United States, including the industrial community of Donora. Of the population of 14,000, 20 died (compared with the expected 2 deaths for the period) and 43% fell ill. Again, the elderly were the most seriously affected. Similar episodes of smog-induced mor- tality in London as early as 1873 and in New York City during the 1950s and 1960s have been described (Amour, 1986~. In all those episodes, the most vulnerable people were the el- derly and those with pre-existing disease of the cardiopulmonary system. Organic mercury compounds, specifically methyl and ethyl mercury salts, are of toxicologic importance because they pass through the blood-brain barrier, accumulate in the brain, and can produce irreversible damage to the central nervous sys- tem. Chronic exposure to these compounds might occur in the workplace or in the general environment. Intense interest in the toxicity of methyl mercury developed in the 1950s, when a neurologic illness, now called Minimata disease,
160 AGING IN TODAY'S ENVIRONMENT appeared In the families of fishermen living around Minimata Bay in Japan. The bay water was contaminated by mercury waste (inorganic and methyl mercury) from a local chemical plant. In- vestigation determined that the illness resulted from the repeated ingestion of large quantities of fish in which the mercury concen- tration was thousands of tunes that in the water (Goyer, 1986~. The cancer experience of the Japanese atom~c-bomb survivors in Hiroshima and Nagasaki has been carefully assessed in rela- tion to the estimates] whol~body doses received by individual survivors. The Life Span Study sample of the Radiation Ef- fects Research Foundation includes 82,000 atomic-bomb survivors and 27,000 nonexposed comparison persons, among whom 19,606 deaths had occurred by 1974 (Beebe, 1979; Beebe et al., 1978~. These studies have been extremely informative about radia- tion-related solid tumors as well as leukemia. For example, they documented the striking difference in the minimal latent periods, with the minimal period for solid tumors being about 10 years, compared with 2 years for leukern~a. Age at time of exposure appears to be a strong determinant of leukemia risk; the greatest absolute risk is experienced by those exposed at ages ~9 and 50 years and over. Leukemia was the first cancer reported in excess among atomic-bomb survivors, but in the most recent followup stud- ies, the later-appear~ng solid tumors appear to be ~ important as leukemia in terms of absolute risk. Most of the excem cancer deaths from solid tumors among the atomic-bomb survivors have occurred in those over 35 at the time of the blast. It appears, therefore, that the radiation effect manifested as lung, breast, and other solid tumors is observed only in people who have reached the age range normally associated with the incidence of cancer. LIFESPAN MODULATION BY DRUG T1lEATMI:NT The effects of toxic agents on aging or aging processes have not been widely studied (Schneider and Reed, 1985~. Some of the best studies carried out so far have concentrated on the effects of drugs that alter oxidative stress. Usually, mean life span and often maximal life span are used as the end points in assessing the effectiveness of experimental interventions in such studies, although in a few cases lipofuscin accumulation and other end points have been used (BaTin, 1982~.
MODEL SYSTEMS FOR EVALUATING TOXIC AGENTS 161 Although several examples of extension of mean life span in mice, in nematodes (Balin, 1982), in Drosophila (Baker et al., 1985), and in house flies (Sohal and Allen, 1986) have been re- ported, the studies have not been highly replicated and have some- times been complicated by design flaws. For instance, treatment often affects only mean and not maximal life span, or treatment might reduce early life trauma and death, but have no effect on aging itself. Studies might also be complicated by effects of the drug treatment on food intake, in that food intake or body weight is significantly decreased by the drug (BaTin, 1982~. No single drug treatment or dietary additive has been reliably shown to extend life span in any organism.
