Longevity, stability and DNA repair

Mechanisms of Ageing and Development, 9 (1979) 203-223 ©Elsevier Sequoia S.A., Lausanne - Printed in the Netherlands

203

LONGEVITY, STABILITY AND DNA REPAIR*

RONALD W. HART, STEVEN M. D'AMBROSIO and KWOKEI J. NG Department of Radiology and Pharmacology, The Ohio State University, College of Medicine, 410 West Tenth Avenue, Columbus, Ohio 43210 [U.S.A.]

SOHAN P. MODAK Visiting Professor in Radiology, The Ohio State University College of Medicine, 410 W. Tenth Avenue, Columbus, Ohio 43210 (U.S.A.] (Received August 10, 1978)

SUMMARY The functional capacity of a cell, tissue, organ, or organism is dependent upon its ability to maintain the stability of its unit components. The higher the differentiated state of the system, the greater the amount of stability required to maintain that state as a function of time. Stability can be achieved via either redundancy or repair. Redundancy while easily achievable in biological systems is both costly and limited by thermodynamic considerations. Repair, in its general sense, has no such limitations. Repair at the cellular and macromolecular level is multiple in its forms and varies as a function of species, tissue, and stage of the cell cycle. The repair of DNA damage is a dynamic process with many components and subcomponents, each interacting with one another in order to achieve a balance between individual stability and evolutionary diversity. Thus, between internal and external factors which damage DNA and the subsequent expression of alterations in the functional stability of DNA lie the multi-functional pathways which attempt to maintain DNA fidelity. A strong correlation between ultra-violet light induced excision or pre-replication repair, as measured by autoradiography and maximum species lifespan has been reported within different strains of the same species, between related species (e.g. Mus musculus and Peromyscus leucopus), between five orders of mammals, and most recently within members of the primate family. As has been demonstrated by the authors and others, differences in excision repair between species and tissues may relate to the turning off of portions of the repair processes during embryogenesis. Regardless of why such correlations exist or the nature of their mechanisms, it is naive to either assert or deny a causal role for DNA repair in longevity assurance systems. For example, while species-related differences in DNA repair may reflect the turning off of such repair processes during fetal development this does not mean that rates of accumulation of DNA damage are not altered by *Based on a paper presented at a specialgroup of Symposia entitled, "Frontiers in AgingResearch", arranged by the program committee of the Biological Sciences Section of the Gerontological Society, San Francisco Meeting, November 18-22, 1977.

204 such changes. Indeed, such a phenomena might well explain the rapid evolution of lifespan within the primates without a concurrent input of new genes.

INTRODUCTION During early morphogenesis, the formation of basic germ layers and the ensuing cell-cell interactions lead to the establishment of organ- or tissue-specific "stem" cell lines. Initially multipotent, progeny become committed towards the expression of an increasingly restricted number of specialized phenotypes. Such a process leading to acquisition by cells of specialized structures and/or functions is called "cell differentiation". Implicit in such a cascade-type event is the progressive loss of the potential for metaplasia or transdetermination, i.e., the stability to be reprogrammed and converted into another phenotype. One of the few well-established exceptions to this, among vertebrates, is the Wolffian lens regeneration system in lentactomized newts where fully differentiated melanotytes from dorsal iris dedifferentiate, proliferate, and then redifferentiate into lens fiber cells [ 1]. The majority of differentiated cell types never divide and the stability of this state seems to be closely related to the potential for cell proliferation and metaplasia. For example, differentiated fibroblasts, lymphocytes, chondrocytes, etc., maintain the ability to re-enter the cell cycle and their progeny redifferentiate into the same basic celt type. While differentiated hemopoietic cells, intestinal epithelium, etc., may not divide, such cell populations are continuously replenished through a pool of proliferating "blast" cells. In the case of hepatocytes, the proliferative potential is expressed upon injury. Terminally differentiated cells lose their proliferative potential more or less irreversibly and can be further subdivided into those which: (a) maintain a transcriptionally active genome (e.g. neurons), (b) those maintaining a transcriptionally inactive, although reactivable, genome (e.g. avian erythrocytes), and (c) those which lose their genome completely while maintaining a stable differentiated state (e.g. lens fibers). DNA is the primary carrier of genetic information and the structural integrity of DNA is a prerequisite for gene expression. Impairment of genome integrity results in either a complete loss or a modification of expression. Thus, the molecular mechanisms involved in repair of damaged portions of DNA and their restitution into functionally intact informational units is fundamental to the maintenance of the genomic integrity [2]. Defective DNA repair could cause an accumulation of lesions or mutations which might be either lethal, lead to an altered phenotype, or neoplastic transformation. In postmitotic cells, such damage may stimulate cell division or impair normal cellular function through a deletion of, or physical alterations in, one or many structural genes. Studies on the integrity of DNA in terminally differentiating and aging cells have revealed that the genome progressively accumulates strand breaks and gaps and has led to the proposal that the integrity of the genome is controlled by DNA repair machinery [2, 3]. Thus, defective repair could result in accumulation .of transcriptionally inactive genes [2]. Considerable

205 evidence has accumulated during the past six years supporting this proposal, and an increasing effort has been directed towards elucidation of various aspects of this phenomenon. The ability of a cell to repair damaged portions of its genome is decisive in maintaining its functional integrity [2]. Repair potential in adult organisms seems to be defined during embryogenesis and cell differentiation [4], and correlates with the lifespan of the organism [3] ; the larger the repair potential, the longer the lifespan. Therefore, it seems that the repair potential plays a key role in defining the integrity of the genome and functional state at the molecular, cellular and organismic level [2]. Sacher and Hart [5] defined two basic paradigms for research on aging and longevity. These are the ontogenetic and the phylogenetic approaches. The former attempts to define the parameters of aging at the molecular and cellular level in cells and tissues from an organism during its lifespan, while the latter seeks to correlate genetic and biochemical parameters to the differences in longevity seen between species. Both of these approaches can be used to examine the role of DNA repair in maintenance of longevity assurance parameters. In this report, we review DNA repair mechanisms controlling genomic integrity in relation to the stability of the differentiated state and discuss the phylogenetic implications of repair potential on longevity.

DNA DAMAGE Structural parameters The DNA molecule consists of a backbone of alternate phosphate and 2-deoxyribose units joined together by phosphodiester bonds at the C-5' and C-3' positions of the sugar moiety. The resultant phosphodiester linkages are susceptible to either chemical or enzymatic hydrolysis. To each deoxyfibose is attached one of predominately four bases (adenine, guanine, cytosine, and thymine) through a N-glycosidic bond. Bases are planar molecules which are relatively insoluble in water. The resulting deoxyribonucleotides, when joined together, form a single-stranded polynucleotide which is partly but not completely randomly coiled in an aqueous environment due to various conformational constraints. The main degrees of rotational freedom for these molecules are limited to the two O-P bonds in the phosphodiester linkage and the glycosidic bond of the base to the sugar moiety. In a native DNA molecule two complementary single-stranded polynucleotides are twisted together in an antiparallel fashion to form a right-handed a-helix. This helix has two grooves, one shallow, one deep. The bases from the two strands are paired in one of the following manners: G-C, C--G, A-T, or T-A. All base pairs have approximately the same size and shape. Within the helix, the base pairs are stacked one on top of another, forming parallel surfaces perpendicular to the helical axis, and also perpendicular to the plane of the deoxyribose. The double-stranded helical DNA under normal physiological conditions has a uniform diameter of about 20 A, with a pitch of 34 .~. Since there are 10 base pairs per turn, each base pair is 3.41 .~ from the next and rotates 36 ° relative to its neighbour.

206 This double-stranded a-helical structure is stabilized under normal cellular conditions by: (a) G-C and A-T base-pairing (hydrogen bonding); and (b) the stacking of the aromatic surfaces of the base pairs in the center of the helix thus exposing the anionic phosphate groups on the outside of the helix to interaction with the cationic species of the aqueous environment. Of these stabilizing interactions, the hydrophobic interaction between the stacked bases is more important than that due to hydrogen bonding between the bases. Double-stranded DNA can be denatured into single-stranded DNA by changes in pH, ionic strength, temperature, and DNA dirnerization or formation of DNA adducts. Native DNA from various organisms denatures at species-specific temperatures depending upon the G-C content of the DNA. The higher the G-C content, the higher the melting temperature (Tin). Functional parameters

The function of DNA is to store and regulate the expression of information required for growth, development, differentiation and cellular maintenance. The stability of this information is directly related to the stability of the cell, tissue and species as a function of time. In its simplest form, the information transfer is accomplished by a series of enzymatically and synthetically regulated steps from DNA to RNA to protein. In actual practice, the situation is far more complex and it can be estimated that at least seventy separate functions must be performed to accomplish the transfer of a single bit of information into a form utilizable by the cell. Further at any single time several hundreds of bits of information must be processed concurrently in a precise sequence and any alteration at any one of these steps will lead to altered cellular function and responsiveness to endogenous and exogenous stress factors. Indeed, several hypotheses on aging are based on one or another of these altered parameters. In other words, the nature and manner of expression of DNA damage will depend on its form, location, repairability and duration at both molecular and cellular level. The consequence of damaged DNA may or may not be mutagenic and/or lethal, and may lead to either repression or depression of particular gene/(s). Since all genetic information is stored in complementary double-stranded DNA, alteration in either the primary, secondary or tertiary structure of DNA can lead to temporal or functional changes in physiological parameters such as cellular metabolism, absorption of nutrients, elimination of metabolic wastes, energy metabolism, replacement of macromolecules, cell division, etc. With this in mind, it becomes of interest to look at DNA damage per se in relation to the aging process and not only as a precursor to mutagenesis. While there is strong evidence suggesting that DNA damage plays a role in carcinogenesis via its mutagenic potential, there is neither evidence for, nor need to, presume that it plays a role in aging via the same pathway. DNA damage

Numerous physical and chemical agents damage cellular DNA in vivo. The induction of such damage may arise due to interaction of cellular DNA with either endogenous or

207 exogenous nucleophilic agents or free radicals. The forms of DNA damage induced vary not only for different agents, but also as a function of the target macromolecules they interact with and the specific atom of interaction on a given target molecule. Thus, interaction of an electrophilic agent with DNA may lead to as many as twenty types of damages including adducts, phosphotriesters, crosslinks, strand-breaks, hydrations, and dimers. The most relevant form of DNA damage with respect to aging would seem to be those induced by endogenous biochemical and physical reactions. These forms of damage however, are much less well understood than those induced by model DNA damaging agents (i.e. ultra-violet light, acetoxy-acetylaminofluorene, dimethylbenzanthracene). It is a reasonable assumption, however, that while the mechanics of DNA repair may vary for these agents, in general, studies with these agents serve to illustrate and elucidate the overall concept of DNA damage and repair. Studies on how these agents interact with cellular DNA and produce distortions via either dimerization [6], substitution at various nucleophilic sites on heterocyclic bases [7, 8], phosphotriesters [9, 10], and/or inter- and intra-strand crosslinks [9], have led to a better understanding of how specific alterations may leadto various biological endpoints. The alkylating agents form a classical group and many of these have been termed radiomimetic [11]. These agents can be divided into two groups: those requiring metabolic activation and those not requiring such activation [12]. In general, the most reactive site on DNA towards methylation is the N7 position of the base guanine (Gua-N7), which lies at a site in the major groove of the DNA helix [13]. However, substitution on GuaN7 appears not to be mutagenic, since the product 7-methylguanine can correctly pair with the DNA base on the opposite strand (cytosine) during DNA replication, thereby not altering the normal nucleotide sequence [14]. While these lesions may not be mutagenic and therefore are presumably non-carcinogenic, they may nevertheless induce functional alterations in cellular physiology. Loveless [15] found that both the strong mutagen and carcinogen methylnitrosourea (MNU) and the weakly mutagenic and carcinogenic alkylating agent methylmethane sulfonate (MMS) predominantly produce a N7-methylguanine product when reacting with cellular DNA. However, he also found that the more potent carcinogen MNU also methylates the guanine O6-position. Subsequent studies confirmed these findings and showed that MNU produced approximately 400 times more O6-guanine product than did MMS [16]. Additional studies have shown that while virtually all potential nucleophilic sites on the bases are susceptible to substitution by alkylating agents; those with low mutagenic potential tend to react primarily at ring nitrogens, whereas those with a high mutagenic potential additionally interact with base oxygens [17]. The alkylation at the O6-position of guanine has now been correlated with mutagenesis [ 15], carcinogenesis [ 18], and transcriptional errors [ 19]. It is interesting to note that while alkylating agents are often deemed as radiomimetic, their major site of attack on DNA may be different from that of ionizing radiation. For example, it now appears that the lethal effects of ionizing radiation may not be due to strand breakage, but rather to thymidine damage [20]. Which of the various forms, if any, of the DNA damage induced by alkylating agents and ionizing radiation is responsible for those age-related physiological processes accelerated by these agents has not yet been determined.

208 In addition to alkylating agents and ionizing radiation, several other classes of chemical agents are known to produce DNA damage including several biological products such as the aflatoxins. These agents represent a class of compound produced by Aspergillus molds. Only in the presence of metabolic activation [21] are these compounds potent hepatocarcinogens [22, 23], and mutagens [24-26] which bind to the N7-position of guanine and yield chromosomal aberrations [27, 28], sister chromatid exchanges (unpublished results), DNA strand-breaks [29], and dominant lethal mutations [30]. A third group of DNA damaging agents which are produced in nature, as well as being byproducts of our industrial society, are the polycyclic aromatic hydrocarbons (PAIl). These have been implicated in various age-related processes including arteriosclerosis, cancer, sister chromatid exchanges and decreased replication potential [12]. Members of this class of agents require metabolic activation and exhibit a wide range of carcinogenic potentials [31, 32]. The polycyclic aromatic hydrocarbons have been shown to bond covalently to DNA in a variety of in vivo, in vitro, and cell-free systems where metabolic [33], chemical [34], or physical [35] activation is possible. While it now appears certain that the biological activation of the PAHs involves the aryl hydrocarbon hydroxylase system of the microsomal membrane, the exact chemical nature of the activated intermediates has been subject to controversy. However, in the last two years, it has become clear that the metabolite responsible for the binding of one member of this diverse group of compounds, benzo(a)pyrene (BP), to cellular DNA is the 7,8-dihydrodiol-9,10oxide of the hydrocarbon [36, 37]. This metabolite is highly mutagenic in both bacterial and mammalian systems [38, 39]. Recently, the elucidation of the structure of the BPRNA reaction product has shown that this hydrocarbon binds to the N2-position of guanine [40, 41 ]. Recent studies on V-79 cells grown in the presence of lethally irradiated rat cells (which metabolize PAH) have shown a strong correlation between mutagenicity and carcinogenicity [38, 42]. These studies support the contention that DNA is the primary target for the carcinogenic action of PAHs and the earlier studies by Brookes and Lawly who found that the carcinogenicity of a particular PAH correlated with the extent to which it binds to DNA, but not to other cellular macromolecules [43]. Additional support for DNA being the primary target for the carcinogenic action of PAH comes from a recent study showing a direct correlation between the tumor-initiating ability of the PAHs in mouse skin and the extent of binding of the compounds to DNA in mouse fibroblasts cultured in vitro for a series of PAHs [44]. Since all of the relationships between carcinogenesis and aging can be viewed as causative in nature [45], and all differences as expressive in nature, it is then a reasonable assumption that the initiating event in both cases may be similar. If, indeed, this is the case, it does not mean that aging is of necessity the result of a clonogenic or mutagenic event, but rather that it may arise from the same event as cancer - namely, DNA damage.

CHROMATIN STRUCTURE Cellular DNA exists not as a naked molecule, but in the form of a complex with histone and non-histone protein [46]. The DNA-histone complex forms a flexible string

209 of closely packed beads or chromatin subunits, also called "nu bodies" or "nucleosomes" [47-50]. The nucleosome itself contains 165-212 base pairs of DNA [54, 55, 58] of which 140 pairs are tightly bound to the historic octamer to form the "core particle" [59-61 ] while the remaining 25-72 base pairs are located in the spacer or linker region. Histone H1 is associated with the spacer [61-64]. A heterogeneity may exist in the repeat lengths of DNA from the same cell lineage at: (a)various stages of differentiation [65] ;(b) within a tissue [66] ; or (c) between different tissues of the same species [58]. Another line of evidence on the existence of repeating chromatin subunits in vivo is derived from terminally differentiating lens fiber cells of chick embryo. In these cells there are discrete size classes of DNA, which are integral multiples of the smallest size class, which appear as fiber cell nuclei undergo natural degeneration in situ [67, 68]. Repeating classes of DNA fragments are also found in radiation-induced cell death [69] and often appear in the cytoplasmic postmitochondrial particle fractions [70, 71] of cells labelled with high concentrations of radioactive precursor [72]. It has been proposed that individual nucleosomes may be composed of symmetrically paired half-nucleosomes each containing one molecule each of H2A, H2B, H3, and H4 histones and 100 base pairs of DNA, so that the two halves might transiently unpair to uncoil the DNA so as to allow genetic readout [73,74]. Several lines of evidence suggest that transcriptionally active regions of chromatin may be structurally distinct from inactive regions [75-83]. While electron microscopic visualization of lampbursh chromosome loops and nucleolar chromatin from amphibians suggests that transcriptionally active regions of chromatin may lack nucleosomes altogether [84], molecular hybridization studies [85] do not support this possibility. The linker-DNA in chromatin is rendered acid-soluble either by the endogenous Ca2÷/Mg2*-dependent endonuclease [51, 52] or by exogenous micrococcal nuclease [53]. In postmitotic differentiated cells, approximately 50% of chromatin-DNA is sensitive to this enzyme [59, 60, 86, 88] with the resistant fraction being associated with the "nucleosome core particle" [59, 60]. In proliferating cell populations as much as 75-80% of chromatin-DNA is hydrolyzed by staphylococcal nuclease [55, 89; Modak and D'Ambrosio, unpublished observations]. It thus seems that the chromatin backbone is not a fixed structure and DNA-histone complexes may be modulated considerably depending on cellular metabolic activity. For example, the histone octamer may slide along DNA [67, 90] and render sequences accessible for DNA replication and transcription. Thus, one may ask whether or not the subunit structure of chromatin controls the location of DNA sites sensitive to various DNA damaging agents. The answer is affirmative for both endogenous and exogenous Ca2÷//Vlg2÷-dependentnucleases in vitro [51, 53] and in vivo [67, 68], very little is known, however, of other chemical and physical agents. Wilkins and Hart [91 ] found that a significant portion of pyrimidine dimers induced by u.v. are masked in nuclei and can be made fully accessible in vitro to u.v.-endonuclease by high salt treatment, thereby showing for the first time preferential DNA repair. Treatment of mouse mammary cells with methylmethane sulfonate and subsequent analysis of the damaged sites indicates that they are non-uniformly distributed or that the repaired regions are distributed non-randomly relative to the nucleosomes [92].

210 In u.v.-irradiated human fibroblasts the initial repair replication seems to occur preferentially in the linker region [93] ; nothing is known, however, concerning the frequency of occurrence of pyrimidine dimers in DNA associated with chromatin subunits. These studies [91--93] indicate that at least the accessibility of the damaged site in DNA is controlled by the chromatin organization but says little about the precise structural parameter involved. A classical case demonstrating the complexity of this issue is found in V-79 cells in which 85-90% of u.v.-induced dimers remain unexcised after six hours [94], although only 20% of chromatin-DNA is associated with staphylococcal nuclease-resistant nucleosome core particles (Modak, D'Ambrosio and Hart, unpublished data). Thus, any model conferring upon the chromatin subunits the role of controlling the accessibility of damaged sites to repair enzymes seems naive unless it takes into account histone-histone, histoneDNA and non-histone-DNA interactions. Thus, the existing data [91-93] emphasize the importance of the chromatin organization and make it clear that a direct extrapolation of results on excision repair in purified DNA is not sufficient to comprehend the situation in vivo.

DNA REPAIR Numerous physical and chemical agents attack cellular DNA in vitro and in vivo yielding at least three forms of DNA damage: (a) structural alterations which include pyrimidine dimers and chemically induced adducts to DNA bases; (b) single-strand resulting from X-irradiation, chemical insult or endonuclease nicking of damaged DNA; and (c) apurinic and apyddiminic sites resulting from chemical binding to purines and pyrimidines, heat or X-irradiation. Figure 1 shows the three general types of DNA repair processes (excision, strandbreak and postreplication) known to take place in mammalian cells following chemical and physical insult. Each of these repair systems have several components which are active in both dividing (excision, strand-break and postreplication) and in non-dividing (excision and strand-break) cells. In excision or prereplication repair, the damage is excised from the parental strand by a number of enzymes. Endonucleases recognizing a specific type of damage (e.g. pyrimidine dimers) or distortions in DNA (e.g. Sl-endonuclease)make single-strand breaks near the damage in the parental strand of DNA. Depending upon the type of damage, exonucleases will remove about 10--200 bases leaving a gap in the parental strand which is filled-in by a polymerase using the opposite DNA strand as a template. A ligase then joins the two strands of DNA together. Excision repair is measured [95] by: (a) assaying for the loss of radiolabeUed DNA-chemical adducts or pyrimidine dimers from DNA; (b) loss of endonuclease sensitive sites in DNA; (c) decrease in the binding of damaged-DNAdirected antibody molecules to DNA [96, 97] ; (d) measuring the amount of unscheduled DNA synthesis using autoradiography; or (e) by determining the number of breaks resulting from the photolysis of DNA containing bromodeoxyuridine incorporated into DNA during repair [98].

211 Dividing Cells

Dividing Ceils Non - dividing Cel Is

Normal DNA

Reactiva'don

Replication "~illi ng Lethal ~1 Error

Prone Rel~ir ~

Strudural Alteration

NO ~Gap"

"Gap" Fqlling

~

OR ~ Single Strand Break

~¢~ %,

"-"~" " Apurinic Site

~

Exonuclease 10 - 200 NLEleotide "Gap"

\\ ~%

M utat i 0 n ( Let hal ~nd°n~le)

Endonuclease Preferential .~! Repair~J ~ v ~ \• "C"~"~'~ ~'a@¢

\ ~/~ ~%~

~"""~ P~y......

~Ligase

Normal DNA

Fig. 1. Diagram of types of DNA repair processes in mammaliancells followingchemicaland physical insult. Single-strand breaks observed in DNA following chemical insult of X-irradiation are repaired and thought to involve several components [99,100] which are not completely understood. Strand-breaks arise as a result [95] of: (a)endonuclease nicking of damaged DNA; (b) the action of OH-; (c) spontaneous breakdown of phosphotriesters; and (d) direct alkylation of bases or sugar moieties. Strand-break repair is measured by: (a) determining the number of such breaks by sedimentation of cellular DNA in alkaline sucrose gradients [ 101 ] ; (b) alkaline elution [ 102 ] ; and (c) electrophoresis in alkaline agarose [ 103 ]. Damage in parental DNA which is not repaired by photoreactivation (photomonomerization of u.v.-induced pyrimidine dimers) or excision repair can enter DNA replication in mitotically active cells. It has been observed that after u.v.-irradiation normal DNA replication is inhibited and that the size of newly synthesized DNA is smaller in treated than in untreated cells [104, 105]. There is evidence to suggest that the daughter DNA contains gaps after chemical and physical treatment which are filled-in and joined together [105, 106]. This process of replicating daughter DNA upon a damaged template has been referred to as postreplication repair. Various models [104-108] by which this process might occur have been discussed and its role in mutation, carcinogenesis and aging are being investigated. Since this repair process involves replicating DNA, its importance in dividing cells is obvious. There appears to be a direct correlation between maximum achievable lifespan and the amount of unscheduled DNA synthesis as measured by autoradiography after u.v.irradiation among different classes [109], as well as within a single class (the primates) of placental mammals [ 110]. There is very good correlation between the amount of excision repair and sensitivity of cells treated with u.v. and AAAF derived from patients having xeroderma pigmentosum [111], Fanconi's anemia [112], and Bloom's syndrome [113].

212 Similar correlations exist with X-rays and alkylating agents for ataxia telangiectasia [114] and progeria [115, 116]. The latter observation, however, has been questioned [117]. Other data [118, 119] suggest that excision repair is an error-free process since cells allowed to excise their damage by holding them in a non-growing state before entering replication are less mutable and transformable than those allowed to enter replication. Further evidence that excision repair is an error-free process comes from the observation [120, 121] that XP cells defective in excision repair are more mutable than normal cells exhibiting a larger amount of repair. These studies imply a role for error-prone repair in mutagenesis and transformation. The mutagenic and carcinogenic potential of any DNA damaging agent is probably initiated when such damage enters the replication machinery as it can in rapidly dividing cells or in cells having low levels of excision repair. Studies with cells derived from another class of XPs which exhibit all the clinical symptoms of sensitivity to sunlight but have normal excision repair, are highly mutable and sensitive to killing by u.v. and AAAF [120]. These cells have also been found to be defective in postreplication repair following u.v. [104] and AAAF [122] treatments. Apparently these XP variants and classical XP cells, low in excision repair, are able to recover or induce an enhanced rate of postreplication repair when first given a small dose of u.v. or AAAF a few hours prior to a larger insult [122]. It is unclear whether enhanced postreplication repair in mammalian cells plays a direct role in u.v. or chemical mutability and carcinogenicity like that observed in bacterial cells. However, it is clear from the cytotoxic and mutagenic data and the correlation of postrepllcation repair with a number of syndromes (clinically observed to be highly prone to environmentally induced cancer and exhibiting certain facets of premature aging) that postreplication repair plays a significant role in mutagenesis, carcinogenesis and possibly aging.

ROLE OF CHROMATININ DNA REPAIR It is generally assumed that the distribution of damaged sites in DNA produced by ionizing radiation or u.v. is random. While this may hold true for naked DNA, there is as yet no evidence on the validity of the above assumption for chromatin-associated DNA. On the other hand, existing data [91--93] show that not all damaged DNA sites in chromatin are uniformly accessible to repair enzymes. When u.v.-irradiated cell nuclei are incubated with u.v.-endonucleases, the accessibility of enzyme-sensitive sites depends on the salt concentration used to expose them; the majority of sites remain unaccessible under low salt conditions but become accessible after high salt treatment [91 ]. Thus, the high salt-sensitive DNA--protein complexes mask u.v.-endonuclease sensitive sites. In cells treated with methylmethane sulfonate [92] or u.v. [93], the unscheduled DNA synthesis occurs preferentially in internucleosomal- or linker-DNA which is readily hydrolyzed by staphylococcal nuclease. However, a number of questions still remain unanswered concerning the differences in the levels of resistant DNA fractions in proliferating and stationary-phase cells. For example, in rapidly dividing cells, 75-80% of chromatin-DNA is

213 staphylococcal nucleas-sensitive [55, 89, Modak and D'Ambrosio, unpublished data, 1978] while this fraction contains only 50% of chromatin-DNA in postmitotic cells from young adults [59, 60, 86-88]. Thus, the structure of chromatin in proliferating cells seems to be comparable to neither stationary phase cells nor to postmitotic tissues. An extreme situation is found in aged mouse liver chromatin containing only 38% of nuclease-sensitive DNA [88]. Furthermore, the effect of inhibitors of semi-conservative DNA replication (e.g. hydroxyurea) on basic structure of chromatin as well as its supraorganization is not known. In the case of u.v. radiation, the formation and consequence of DNA--protein crosslinks on the accessibility of pyrimidine dimers to repair enzymes on one hand, and the relative mobility of the histone octamer along the DNA on the other, is not understood at all. Considering the strength of binding between the histone octamer and DNA [123, 124] as well as the ratio of protein to DNAin core particles [56, 57] as compared to the linker region, the conformational restraints excercised by formation of proteinDNA crosslinks cannot easily be ignored. Finally, it is known that polyamines affect the chromatin structure [125] and their role in controlling the accessibility of damaged sites in chromatin-DNA to repair enzymes has never been investigated. Perhaps the most overwhelming consideration is that, like the ribosome, DNA replication repair and enzyme molecules represent large complexes, and must require at least a temporary weakening of the DNA-histone complex, if not its complete dissociation or relative displacement in order to render template DNA accessible to them.

TERMINALCELL DIFFERENTIATIONAND AGING A number of studies in the early 1970s suggested a considerable number of similarities in the metabolic events affecting DNA in terminally differentiating cells, both in embryos and adults, and postmitotic aging cells. These studies and the more recent work in this field implicate the chromatin subunit organization as well as DNA repair machinery in controlling the integrity of the genome.

Genome integrity in postmitotic cells Exogenous DNA polymerase-a and terminal deoxynucleotidyl transferase catalyze incorporation of radiolabelled deoxyribonucleotides into DNA in fixed cell nuclei using endogenous DNA primers [2, 126-133]. The incorporation as revealed by radioautography and quantified by grain counting takes into account various stereomorphometric parameters affecting the radioautographic efficiency [ 127-130]. DNA polymerase utilizes denatured DNA with bound 3'-OH ends facing the single-stranded regions as template [ 134], while terminal transferase requires free 3'-OH ends on single-stranded DNA stretches as initiators [135]. It is possible to identify 3'-OH ends in DNA of unknown physical conformation using both of the aforecited enzymatic reactions. Using these assay systems, it was found that during the differentiation of lens fiber cells and vaginal keratinizing epithelia, nuclear DNA accumulated free 3'-OH ends, indicative of single-strand breaks [2, 128-131]. Sedimentation studies using alkaline sucrose gradients [136, 137] and two-

214 dimensional electrophoresis in native and alkaline agarose gels [138] have confirmed the above findings. The template activity for DNA polymerase-a of DNA in fixed cell nuclei of cerebral cortex neurons, liver kupffer cells and cardiac muscle was examined and found to increase with age [132, 133], and it was suggested that DNA in these cells accumulates gaps and strand-breaks during the aging process [2, 132, 133]. Alkaline sucrose gradient sedimentation studies on DNA from rat liver [139] and dog neurons and retinal photoreceptors [140] have confirmed the above findings by demonstrating that the single-strand molecular weight of DNA in these cells decreases with age. It should, however, be noted that in alkaline sucrose gradients not only is DNA denatured, but also alkali-labile bonds such as apurinic acid residues are broken and these may partly contribute to the decreased single-strand molecular weight. In cytochemical procedures, however, the denaturation was carried out in 0.01 N HC1 which does not seem to be able to generate additional strand breaks as analyzed on either native or alkaline agarose gels [ 141 ]. In another study, Karran and Ormerod [142] found that DNA in terminally differentiating chicken erythrocytes and rat muscle accumulates strand breaks. Viewed together, these studies indicate that the integrity of the genome is affected in both terminally differentiating cells and aging postmitotic cells in a similar manner, and point towards a common mechanism involving defective repair of strand breaks. It should be emphasized that nothing is known of the possible accumulation of apurinic sites and adducts during the aging process and this aspect deserves attention in the near future. Clearly, the unrepaired lesions in DNA will constitute obstacles for polymerases involved in both replication and transcription, and lead to a progressive inactivation of the genome [2]. Thus, the conservation of effective DNA repair machinery represents the most important single element controlling the genome integrity [2].

Repair potential and defects The status of DNA repair systems have been analyzed for a number of cell types. During in vitro cellular aging, excision repair or unscheduled DNA synthesis seems to decrease in late passage cells [109, 143-146]. During the terminal differentiation of skeletal muscle, unscheduled DNA synthesis decreases and disappears [147-149]. U.v.irradiated lens fibers are unable to carry out unscheduled DNA synthesis while lens epithelial cells do so [ 150]. The rejoining of single-strand breaks after X-irradiation has been found to be absent in lens fiber cells, but present in lens epithelial cells [ 137]. The molecular species of DNA polymerases in eukaryotes responsible for repair replication have yet to be fully identified. However, DNA polymerase-a activity decreases and disappears in differentiating skeletal muscle [151]. Characterization of DNA polymerase-/3 in aging mouse spleen suggests that this enzyme loses its fidelity of complimentary strand synthesis with age [152]. An age-dependent decrease in the fidelity of DNA polymerase-~ was found to occur during cellular aging [153]. As predicted before [2], it appears that terminal cell differentiation and the aging process are accompanied by an appearance of defective repair systems. A great deal of enzymological and biochemical studies are necessary to understand the precise nature of the defects which occur in repair enzymes as a function of age.

215 Chromatin structure

In the earlier part of this review, we have described the important features of eukaryotic chromatin and discussed its possible role in control of replication, transcription, and repair. It now appears certain that DNA in transcriptionally-active and repair-active fractions of chromatin is readily accessible to DNA hydrolyzing enzymes. Thus, the template-active and repair-active fractions of chromatin seem to be structurally distinct from inactive regions [73-83, 91-93 ]. In attempts to investigate the relationship between alterations in cellular macromolecules and aging, several authors have analyzed the thermal melting behavior [154156] and DNA template activity for exogenous RNA polymerase [154, 158-160] of chromatin in aging tissues with conflicting results. Chromatin Tm increases as a function of age in bovine thymus and rat liver [154, 158], while it has been reported to first decrease during the maturation and then to increase back to the original level in aging mouse brain [156, 157]. On the other hand, with the exception of Samis and Wulff [160], all authors [154, 158, 159] find an age-dependent decrease in the chromatin template activity for exogenous RNA polymerase. Complexity measurements suggest that the number of different types of RNA sequences expressed in mouse tissues decrease with age[ 161 ]. Taken together, these studies indicate that the chromatin structure undergoes discrete changes during the aging process. Relative to this, Modak et al. [88] found that the DNA repeat size remains unchanged as a function of time although in old mice (28-33 months) a considerable heterogeneity appears in the digestion products with staphylococcal nuclease. Perhaps the most significant finding is related to the observation that the proportion of staphylococcal nuclease-sensitive DNA in mouse liver chromatin decreases from 50% (1.75-18 months) to 38% at 28-33 months of age so that the actual proportion of DNA organized into chromatin subunits increases from 70% at 18 months to 92% in very late age [88]. Although these differences suggest that the ratio histone:DNA should increase, this has not been found to be the case [90, 159]. Alternate explanations include the possibility that there exists a free pool of histones in mouse liver cells, or that at early and mid-ages, there exist half-nucleosome-like structures similar to those postulated by Weintraub and collaborators [73, 74] containing DNA which is fully sensitive to staphylococcal nuclease. In the latter case, the half-nucleosome pairs may reassociate in late age to form full nucleosomes, thus rendering resistance of DNA to the staphylococcal nuclease. So far, it has not been possible to ascertain whether non-histone proteins confer a nucleaseresistant property upon chromatin-DNA, but this possibility cannot be excluded at present. Non-histone proteins play an important role in regulation of gene expression [46], but nothing is known of their involvement in age-dependent genome inactivation. Medvedev et al. [162, 163] have recently found that histone H1 is modified in old cells. Two-dimensional electrophoretic analysis of histones from mouse liver chromatin suggest that significant histone modifications may appear at 28-33 months of age [ 164]. Histone modifications can be expected to affect the conformation of chromatin, e.g., an increased charge on H1 would result in condensation of the chromatin while appearance of H4 with greater charge would tend to increase histone:histone interactions while at the same time decreasing DNA:histone interactions [55, 164]. Thus the present evidence

216 suggests that significant changes occur in chromatin structure as a function of age. This may involve [88] an increased concentration of nucleosomes which may lead to a decreased mobility of histone octamers along DNA with a subsequent decrease in the accessibility of DNA to repair enzymes. However, detailed analysis of chromatin structure in a variety of organ systems and comparative-evolutionary model systems is necessary to understand the exact molecular basis of changes in DNA:histone and histone:laistone interactions.

Repair potential and longevity Senescence can be defined as the loss of homeokinesis at several levels of biological organization [165]. Molecular and biological studies suggest that the time-dependent changes in the levels of homeokinesis may depend upon the ability of a given system to maintain the fidelity of its DNA. Conversely, any alteration in the structure of DNA would be expected to lead to alteration in cellular function and thus the ability of that system to respond to external and internal stress factors. Since there exist numerous forms of DNA damage and DNA repair systems, it is expected that either a single type of DNA damage or a lack of any individual type or repair system would mimic certain aspects of the aging process, but none of these would individually mimic all aspects of senescence [166]. Indeed, this contention is supported by the situation for DNA repair defective human syndromes such as xeroderma pigmentosum, ataxia telangectasia, progeria, Bloom's syndrome, Fanconi's anemia, etc., each of which exhibit some aspects of the aging process. l.~kewise, different DNA-damaging agents bring forth expression of different aspects of senescence. The situation is obviously far more complex since individual cell types exhibit differential sensitivity to different DNA-damaging agents. Thus, at the organismic level, an examination of all types of DNA repair systems, during the in vivo aging process, is necessary to define the role of DNA repair in relationship to longevity. While such detailed studies are still missing, the repair potential of u.v.-irradiated fibroblasts from a variety of mammalian species does reveal a direct correlation between repair and maximum achievable lifespan [ 109, 110]. Thus, the repair potential may be one of the decisive elements in controlling lifespan. However, the repair potential itself does not seem to be constant throughout the lifespan [143,167]. In all mammalian species studied, the level or repair activity is high in early embryo cells and, towards the end of the gestation period, it decreases to attain apparent steady-state levels differing among different species [4], and apparently related to their maximum achievable lifespan. The precise significance of this finding is unclear; however, it may be related to a decreased regulatory potential towards the end of embryogenesis and fetal life. It would be interesting to know in lower vertebrates showing a higher frequency of adult tissue-metaplasia, whether the relative levels of repair activity show a much smaller difference among early and late embryonic and adult stages. In any event, studies analyzing the repair potential are generally being carried out on dividing cells and only little can be said so far of postmitotic cells which undergo terminal differentiation. It is our feeling that the repair potential of individual cell types may be related to their cell proliferation potential, and a loss of either would be rant-

217 amount to senescence [ 168], i.e., inability to maintain the integrity of the genome and a consequent cell death [2]. One of the most spectacular demonstrations of the dependence of the lifespan on the integrity of DNA has been recently presented for paramecia [169]. Smith-Sonnerborn showed that photoreactivation subsequent to u.v.-induction of pyrimidine dimers resulted in a prolongation of the lifespan, while u.v.-irradiation alone and subsequent growth in the dark resulted in a lifespan shortening. These results constitute a formal proof that DNA damage and potentially repair activity is directly related to lifespan.

ACKNOWLEDGEMENT This work was supported in part by US Environmental Protection Agency Grant No. R-80533701-0.

REFERENCES 1 T. Yamanda, Control mechanisms in cell-type conversion in newt lens regeneration, in A. Wolsky (ed.),Monographs in DevelopmentalBiology, Vol. 13, Karger, Basel, 1977, pp. 1-126. 2 S. P. Modak, A model for transcription control in terminally differentiating lens fiber cells, in R. Harris, P. Allin and D. Viza (eds.), Cell Differentiation, Copenhagen, Munksgaard, 1972, pp. 439-442. 3 R. W. Hart and R. B. Setlow, Correlation between deoxyribonucleic acid excision-repair and lifespan in a number of mammalian species, Proc. Nat. Acad. ScL U.S.A., 71 (1974) 2169--2173. 4 L. Peleg, E. Raz and R. Ben-Ishai, Changing capacity for DNA excision-repair in mouse embryonic cells in vitro, Exp. Cell Res., 104 (1977) 301-307. 5 G. A. Sacher and R. W. Hart, Longevity, aging and comparative cellular and molecular biology of the house mouse, Mus musculus, and the white-footed mouse, Peromyscus leucopus, in D. Bergsma and D. Harrison (eds.), Birth Defects: OriginalArticle Series, Vol. 14, Alan R. Liss, New York, 1978, pp. 71-96. 6 R. B. Setlow and W. L. Carrier, The disappearance of thymine dimers from DNA: an error-correcting mechanism, Proc. Nat. A cad. ScL U.S.A., 51 (1964) 226-231. 7 C. C. Irving, Interactions of chemical carcinogens with DNA, in F. Coulston, F. Korte and M. Goto (eds.), New Methods in Environmental Chemistry and Toxicology, Inter. Acad. Print. Co., Tokyo, 1973, pp. 119-134. 8 J. W. Drake and R. H. Baitz, The biochemistry of mutagenesis, A. Rev. Biochem., 45 (1976) 11-37. 9 P. D. Lawley, Effect of some chemical mutagens and carcinogens on nucleic acids, Prog. Nucleic Acid Res. MoL Biol., 5 (1966) 89-131. 10 P. D. Lawley, Reaction of N-methyl-N-nitrosourea (MNU) with a2Pqabelled DNA: Evidence for formation of phosphotriesters, Chem-Biol. Interact., 7 (1973) 127-130. 11 R. W. Hart, R. E. G~son, J. D. Chapman, A. P. Reuvers, B. K. Sinha, R. K. Griffith and D. T. Witiak, A radioprotective stereostructure--activity study of c/s- and trans-2-mercapto-cyclobutylamine analogs and homologs of 2-mercaptocyclobutylamine, J. Med. Chem., 18 (1975) 323-331. 12 R. W. Hart, K. Y. Hall and F. B. Daniel, DNA repair and mutagenesis in mammalian cells, Photochem. Photobiol., 28 (1978) 131-155. 13 B. Reiner and S. Zamerthof, Studies on the chemically reactive groups of deoxyribonucleic acids, J. Biol. Chem., 228 (1957) 475--476. 14 D. B. Ludlum, Alkylated polycytidylic acid templates for RNA polymerase, Biochim. Biophys. Acta, 213 (1970) 142-148.

218 15 A. Loveless, Possible relevance of O6-alkylation of deoxyguanosine to the mutagenicity and carcinogenicity of nitrosamines and nitrosamides, Nature, 223 (1969) 206-207. 16 P. D. Lawley, Alkylation of nucleic acids and mutagenesis, in L. Prakash, F. Sherman, M. Miller, C. Lawrence and H. Taber (eds.), Molecular and Environmental Aspects of Muragenesis, C. C. Thomas, Springfield, Illinois, 1974, pp. 17-33. 17 B. Singer, Sites in nucleic acids reacting with alkylating agents of differing carcinogenicity or mutagenicity, J. Toxicol. Environ. Health, 2 (1977) 1279-1295. 18 R. Goth and M. F. Rajewsky, Molecular and cellular mechanisms associated with pulse-carcinogenesis in the rat nervous system by ethyinitrosourea: ethylation of nucleic acids and elimination rates of ethylated bases from DNA of different tissues, Z. Krebsforsch., 82 (1974) 37--64. 19 L. L. Gerchman and D. B. Ludlum, The properties of O6-methylguanine in template for RNA polymerase, Biochim. Biophys. Acta, 308 (1973) 310--316. 20 M. C. Paterson, B. P. Smith, P. H. M. Lohman, A. K. Anderson and L. Fishman, Defective excision repair of ~,-ray-damaged DNA in human (ataxia telangiectasia) fibroblasts, Nature, 260 (1976) 444-447. 21 J. M. Essigmann, R. G. Croy, A. Nadzan, W. R. Busby, Jr., V. N. Reinhold, G. Buchi and G. N. Wogan, Structural identity of the major DNA adduct formed by aflatoxin B in vitro, Proc. Nat. Acad. Sci. U.S.A., 74 (1977) 1870-1874. 22 G. N. Wogan, Aflatoxin carcinogenesis, Methods Cancer Res., 7 (1974) 309-345. 23 T. C. Campbell and J. R. Hayes, The role of afiatoxin metabolism and its toxic lesion, Toxicol. Appl. Pharmacol., 35 (1976) 199-222. 24 V. M. Maher and W. C. Summers, Mutagenic action of aflatoxin B1 on transforming DNA and inhibition of DNA template activation in vitro, Nature, 225 (1970) 68-70. 25 T.M. Ong, Aflatoxin mutagenesis, Mutat. Res., 32 (1975) 35-53. 26 D. F. Krahn and C. Heidelberger, Liver homogenate-mediated mutagenesis in Chinese hamster V79 cells by polycyclic aromatic hydrocarbons and aflatoxins, Murat. Res., 46 (1977) 27-44. 27 A. Velazquez, C. DeNava, R. Coutino and I. Pulido, The relationship between gene and chromosome mutations in cultured Chinese hamster cells exposed to aflatoxin B1, Murat. Res., 21 (1973) 241-242. 28 D. A. Dolimpio, C. Jacobson and M. Legator, Effect of aflatoxin on human leukocytes, Proc. Exp. Biol.,Meal., 127 (1968) 559-562. 29 M. Umeda, T. Yamamoto and M. Saito, DNA strand breakage of HeLa cells induced by several mycotoxins, Japan. J. Exp. Med., 42 (1972) 527-535. 30 S. S. Epstein and H. Schafner, Chemical mutagens in the human environment, Nature, 219 (1968) 385-387. 31 C. Heidelberger, The need for additional alkylating agents and antimetabolites, Cancer Res., 29 (1969) 2435-2442. 32 C. Heidelberger, Studies on the cellular and molecular mechanisms of hydrocarbon carcinogenesis, Eur. J. Cancer, 6 (1970) 161-172. 33 H. V. Gelboin, A mierosome-dependent binding of benzo(a)pyrene to DNA, Cancer Res., 29 (1969) 1272-1276. 34 C. E. Morreal, T. L. Dao, K. Eskins, C. L. King and J. Dienstag, Peroxide induced binding of hydrocarbons to DNA, Biochim. Biophys. Acra, 169 (1968) 224-229. 35 G. M. Blackburn, A. J. Flavell, P. E. Taussig and J. P. Wilt, Binding of 7,12-dimethylbenz(a)anthracene to DNA investigated by tritium displacement, J. Chem. Soc. Chem. Commun., 215 (1975) 358-359. 36 P. Sims, P. L. Grover, A. Swaisland, K. Pal and A. Hewer, Metabolic activation of benzo(a)pyrene proceeds by a diol-epoxide, Nature, 252 (1974) 326-328. 37 M. R. Osborne, M. H. Thompson, E. M. Tarmy, F. A. Beland, R. G. Harvey and P. Brookes, The reaction of 7,8-dihydro-7,8-dihydroxybenzo(a)pyrene-9,10-oxide with DNA in relation to the benzo(a)pyrene-DNA products isolated from cells, Chem-Biol. Interact., 13 (1976) 343-348. 38 E. Huberman and L. Sachs, Mutability of different genetic loci in mammalian cells by metabolically activated carcinogenic polycyclic hydrocarbons, Proc. Nat. Acad. Sci. U.S.A., 73 (1976) 188-192. 39 P. G. Wislocki, A. W. Wood, R. L. Chang, ~ . Levin, H. Yagi, O. Hernandez, D. M. Jerina and A. H. Conney, High mutagenicity and toxicity of a diol epoxide derived from benzo(a)pyrene, Biochem. Biophys. Res. Commun., 68 (1976) 1006-1013.

219 40 A. M. Jeffrey, K. W. Jennette, S. H. Blobstein, I. B. Weinstein, F. A. Beland, R. G. Harvey, H. Kasai, I. Miura and K. Nakanishi, Benzo(a)pyrene-nucleic acid derivative found in vitro: structure of benzo(a)pyrenetetrahydrodiol epoxide-guanosine adduct, Z Am. Chem. Soc., 98 (1976) 57145715. 41 A. M. Jeffrey, S. H. Blobstein, I. B. Weinstein, F. A. Beland, R. G. Harvey, H. Kasai and K. Nakanishi, Structure of 7,12-dimethylbenz(a)anthracene-guanosine adducts, Proc. Nat. Acad. ScL U.S.A., 73 (1976) 2311-2315. 42 E. Huberman and L. Sachs, Cell-mediated mutagenesis of mammalian ceils with chemical carcinogens, Int. J. Cancer, 13 (1974) 326-333. 43 P. Brookes and P. D. Lawley, Evidence for the binding of polynuclear aromatic hydrocarbons to the nucleic acids of mouse skin: relation between carcinogenic power of hydrocarbons and their binding to DNA, Nature, 202 (1974) 781-784. 44 S. G. Buty, S. Thompson and T. J. Slaga, The role of epidermal aryl hydrocarbon hydroxylase in the covalent binding of polycyclic hydrocarbon to DNA and its relationship to tumor initiation, Biochem. Biophys. Res. Commun., 70 (1976) 1102-1108. 45 H. C. Pitot, Carcinogenesis and aging - two related phenomena? A review, Am. J. PathoL, 87 (1977) 444-472. 46 A. Ruiz-Carrillo, L. J. Wangh and V. G. Allfrey, Processing of newly synthesized histone molecules, Science, 190 (1975) 117-128. 47 A. L. Olins and D. E. Olins, Spheroid chromatin units (u bodies), J. CellBiol., 59 (1973) 252a. 48 A. L. Olins and D. E. Olins, Spheroid chromatin units (v bodies), Science, 183 (1974) 330-332. 49 C. L. F. Woodcock, Ultrastructure of inactive chromatin, J. CellBiol., 59 (1973) 368a. 50 P. Oudet, M. Gross-Bellard and P. Chambon, Electron microscopic and biochemical evidence that chromatin structure is a repeating unit, Cell, 4 (1975) 281-300. 51 D. R. Hewish and L. A. Burgoyne, Chromatin sub-structure: the digestion of chromatin DNA at regularly spaced sites by a nuclear deoxyribonuclease, Biochem. Biophys. Res. Commun., 52 (1973) 504-510. 52 L. A. Burgoyne, D. R. Hewish and J. Mobbs, Mammalian chromatin substructure studies with the calcium-magnesium endonuclease and two-dimensional polyacrylamide-gel electrophoresis, Biochem. J., 143 (1974) 67-72. 53 M. Noll, Subunit structure of chromatin, Nature, 251 (1974) 249-251. 54 J. O. Thomas, Aspects of the structure of chromatin, in E. M. Bradbury and K. Javaherian (eds.), The Organization and Expression of Eukaryotic Genome, Academic Press, New York, 1977, pp. 83-88. 55 E. M. Bradbury, Histone interactions and chromatin structure, in E. M. Bradbury and K. Javaherian (eds.), The Organization and Expression of Eukaryotic Genome, Academic Press, New York, 1977, pp. 99-123. 56 R. D. Kornberg, Chromatin structure: a repeating unit ofhistones and DNA, Science, 184 (1974) 868-871. 57 R. D. Kornberg and J. O. Thomas, Chromatin structure oligomers of the histones, Science, 184 (1974) 865-868. 58 N. R. Morris, A comparison of the structure of chick erythrocyte and chick liver chromatin, Cell, (1976) 627--632. 59 B. Sollner-Webb and G. Felsenfeld, A comparison of the digestion of nuclei and chromatin by staphylococcal nuclease, Biochemistry, 14 (1975) 2915-2920. 60 R. Axel, Cleavage of DNA in nuclei and chromatin with staphylococcal nuclease, Biochemistry, 14 (1975) 2921-2925. 61 B. R. Shaw, T. M. Herman, R. T. Kovacic, G. S. Beaudrau and K. E. van Holde, Analysis of subunit organization in chicken erythrocyte chromatin, Proc. Nat. Acad. ScL U.S.A., 73 (1976) 505-509. 62 J. P. Baldwin, P. G. Bosely, E. M. Bradbury and K. Ibel, The subunit structure of the eukaryotic chromosome, Nature, 253 (1975) 245-249. 63 A. J. Varshavsky, V. V. Bakayev and G. P. Georgiev, Heterogeneity of chromatin subunits in vitro and location of histone H 1, Nucleic Acids Res., 3 (1976) 477-492. 64 J.P. Whitlock, Jr. and R. T. Simpson, Removal of histone H1 exposes a fifty base pair DNA segment between nucleosomes, Biochemistry, 15 (1976) 3307-3314.

220 65 D. Lohr, J. Corden, K. Tatchell, R. T. Kovacic and K. E. van Holde, Comparative subunit structure of HeLa, yeast and chicken erythrocyte chromatin, Proc. Nat. Acad. Sci. U.S.A., 74 (1977) 79-83. 66 R. D. Todd and W. T. Garrard, Two-dimensional electrophoretic analysis of polynucleosomes, J. Biol. Chem., 252 (1977) 4729-4738. 67 S.W. Appleby and S. P. Modak, DNA degradation in terminally differentiating lens fiber cells from chick embryos, Proc. Nat. Acad. Sci. U.S.A., 74 (1977) 5579-5583. 68 S. P. Modak, D.W. Appleby and C. Unger-Ullmann, Digestion in vivo of chromatin from terminally differentiating chick lens fiber cells, Experientia, 34 (1978) 57 (abstract). 69 M. Skala, J. Matyasova and M. Cejkova, DNA in chromatin of irradiated lymphoid tissues degraded in vivo into regular fragments, FEBS Lett., 72 (1976) 271-274. 70 R. A. Williamson, Properties of rapidly labelled deoxyribonucleic acid fragments isolated from the cytoplasm of primary cultures of embryonic mouse liver cells, J. Mol. Biol., 51 (1970) 157-168. 71 S. P. Modak, D. Commelin, L. Grosset, M. T. Imaizumi, M. Monnat and K. Scherrer, DNA synthesis in circulating erythroblasts of anemic duck, Eur. J. Biochem., 60 (1975) 407-421. 72 S. P. Modak, D. W. Appleby and M. Chappuis, Cytoplasmic informational DNA: Fact or fantasy, J. CeUBiol., 70 (1976) 140a. 73 H. Weintraub, A. Worcel and B. Alberts, A model for chromatin based upon two symmetrically paired half-nucleosomes, Cell, 9 (1976) 409-417. 74 B. Alberts, A. Worcel and H. Weintraub, On the biological implications of chromatin structure, in E. M. Bradbury and K. Javaherian (eds.), The Organization and Expression ofEukaryotic Genome, Academic Press, New York, 1977, pp. 163-191. 75 R. Axel, H. Cedar and G. Felsenfeld, Synthesis of globin dbonucleic acid from duck-reticulocyte chromatin in vitro, Proc. Nat. Acad. ScL U.S.A., 70 (1973) 2029-2032. 76 R. S. Gilmour and J. Paul, Tissue-specific transcription of the globin gene in isolated chromatin, Proc. Nat. Acad. Sci. U.S.A., 70 (1973) 3440-3442. 77 A. Steggles, G. Wilson, J. Kantor, D. Picciano, A. Falvry and W. F. Anderson, Cell free transcription of mammalian chromatin: transcription of globin messenger RNA sequences from bonemarrow chromatin with mammalian RNA polymerase, Proc, Nat. Acad. Sci. U.S.A., 71 (1974) 1219-1222. 78 M. J.acquet, A. Groner, G. Monroy and J. Hurwitz, The in vitro synthesis of avian myeloblastosis viral RNA sequences, Proc. Nat. Acad. Sci. U.S.A., 71 (1974) 3045-3049. 79 G. S. Stein, W. Park, C. L. Thrall, R. J. Mans and J. L. Stein, Regulation of cell cycle stage-specific transcription of histone genes from chromatin by non-histone chromosomal proteins, Nature, 257 (1975) 764-767. 80 J. F. Chiu, Y. H. Tsai, K. Sakuma and L. Hnilica, Regulation of in vitro mRNA: transcription by a fraction of chromosomal proteins, J. Biol. Chem., 250 (1975) 9431-9433. 81 S. E. Harris, R. J. Schwartz, M. J. Tsai, B.W. O'Malley and A. R. Roy,Effect of estrogen on gene expression in the chick oviduct, Z Biol. Chem., 251 (1976) 524-529. 82 H. Weintraub and M. Groudine, Chromosomal subunits in active genes have an altered conformation, Science, 193 (1976) 848-856. 83 A. Garel and R. Axel, Selective digestion of transcriptionally active ovalbumin genes from oviduct nudei, Proc. Nat. Acad. Sci. U.S.A., 73 (1976) 3966-3970. 84 W. W. Frank, U. Scheer, M. F. Trendelenburg, H. Spring and H. Zentgraf, Absence of nucleosomes in transcriptionally active chromatin, Cytobioiogie, 13 (1976) 401-434. 85 J. M. Gottsfeld and D. A. Melton, The length of nucleosome-associated DNA is the same in both transcribed and non-transcribed regions of chromatin, Nature, 273 (1978) 317-319. 86 R. J. Clark and G. Felsonfeld, Structure of chromatin, Nature (New Biol.), 229 (1971) 101-106. 87 R. J. Clark and G. Felsonfeld, Association of arginine-rich histories with GC-rich regions of DNA, Nature (New Biol.), 240 (1972) 226-229. 88 S. P. Modak, C. Goner, C. Unger-Ullman and M. Chappuis, Experientia, 34 (1978) 57 (abstract). 89 J. L. Compton, R. Hancock, P. Oudet and P. Chambon, Biochemical and electron-microscopic evidence that the subunit structure of Chinese-hamster-ovary interphase chromatin is conserved in mitotic chromosomes, Eur. J. Biochem., 70 (1976) 555-568. 90 P. Beard, Positional stability of nucleosomgs on DNA, Experientia, 33 (1977) 812 (abstract). 91 R. J. Wilkins and R. W. Hart, Preferential DNA repair in human cells, Nature (New Biol.), 247 (1973) 35-36.

221 92 W. J. Bodell, Non-uniform distribution of DNA repair in chromatin after treatment with methylmethane sulfonate, Nucleic Acid Res., 4 (1977) 2619-2628. 93 J. E. Cleaver, Nucleosome structure controls rates of excision repair in DNA of human cells, Nature, 270 (1977) 451--453. 94 S M. D'Ambrosio and R. B. Seflow, On the presence of UV-endonuclease sensitive sites in mammalian DNA, in P. C. Hanawalt and E. Freidberg (eds.), DNA Repair Mechanisms, Academic Press, New York, 1978, in press. 95 R. W. Hart, S. Hays, D. Brash, F. B. Daniel, M. T. Davis and N. J. Lewis,In vitro assessment and mechanism of action of environmental pollutants, Ann. N.Y. Acad. Sci., 298 (1977) 141-158. 96 J. J. Cornelius, J. Rommelaere, J. Urbain and M. Erreru, A sensitive method for measuring pyrimidine dimers in situ, Photochem. PhotobioL, 26 (1977) 241-246. 97 M. C. Poirier, S. H. Yuspa, I. B. Weinstein and S. Blobstein, Detection of carcinogen-DNA adducts by radioimmunoassay, Nature, 270 (1977) 186-188. 98 J. D. Regan, R. B. Setlow and R. D. Ley, Normal and defective repair of damaged DNA in human cells: a sensitive assay utilizing the photolysis of bromodeoxyuridine, Prec. Nat. Acad. Sci. U.S.A., 68 (1971) 708-712. 99 C. D. Town, K. C. Smith and H. S. Kaplan, DNA polymerase required for rapid repair of X-rayinduced DNA strand-breaks in rive, Science, 172 (1971) 851-853. 100 T. Lindahl and B. Nyberg, Rate of depurination of native deoxyribonucleic acid, Biochemistry, 11 (1972) 3610-3618. 101 M. M. Elkind and C. Kamper, Two forms of repair of DNA in mammalian cells following irradiation, Biophys. £, 10 (1970) 237-245. 102 K. Kohn, L. Erickson, R. Ewig and C. A. Friedman, Fractionation of DNA from mammalian cells by alkaline elution, Biochemistry, 15 (1976) 4629-4637. 103 S. M. D'Ambrosio, S. P. Modak, D. E. Brash and R. W. Hart, Detection of DNA strand-breaks and UV-endonuclease sensitive sites in X-ray and UV-irradiated cells by alkaline agarose gel electrophoresis, J. CellBioL, 79 (1978) 145a. 104 A. R. Lehmann, S. Kirk-Bell, C. F. Arlett, M. C. Paterson, P. H. M. Lehman, E. A. DeWeerdKastelein and D. Bootsma, Xeroderma pigmentosum cells with normal levels of excision repair have a defect in DNA synthesis after UVhrradiation, Prec. Nat. Aead. Sci. U.S.A., 72 (1975) 219-223. 105 A. R. Lehmann, Postreplication repair of DNA in mammalian cells, Life ScL, 15 (1974) 2 0 0 5 2016. 106 R. B. Setlow and S. M. D'Ambrosio, Size of putative gaps in postreplication repair, Biophys. J., 21 (1978) 94a. 107 N. P. Higgins, K. Koto and B. Strauss, A model for replication repair in mammalian cells, £ MoL Biol., 101 (1976) 417-425. 108 Y. Fujiwara and M. Tatsumi, Replicative bypass repair of ultraviolet damage to DNA of mammalian cells: caffeine sensitive and resistant mechanism, Mutat. Res., 37 (1976) 81-110. 109 R. W. Hart and R. B. Setlow, DNA repair and lifespan of mammals, in P. C. Hanawalt and R. B. Setlow (eds.), Molecular Mechanisms for Repair of DNA, Part B, Plenum Press, New York, 1975, pp. 801-804. 110 K. Hall, C. Albdghtson and R. W. Hart, A direct relationship among primates between maximum lifespan and DNA repair, Abstr. 11th Int. Congr. Gerontology, Tokyo, 1978. 111 J. E. Cleaver and D. Bootsma, Xeroderma pigmentosum: biochemical and genetic characteristics, Ann. Res. Genet., 9 (1975) 19-38. 112 P. K. Peon, J. W. Parker and R. L. O'Brien, Defect in DNA repair in Faconi's anemia,Abstr. Prec. Am. Ass. Cancer Res., 15 (1974) 19. 113 F. Giannelli, P. F. Benson, S. A. Pawsey and P. Palari, UV light sensitive and delayed DNA-chain maturation in Bloom's syndrome, Nature, 265 (1977) 466--469. 114 M. C. Patterson and B. P. Smith, Ataxia talangiectasia. A X-ray analogue of xeroderma pigmentosum, Murat. Res., 46 (1977) 148 (abstract). 115 J. Epstein, J. R. Williams and J. B. Little, Deficient DNA repair in human progeroid cells, Prec. Nat. Acad. Sci. U.S.A., 70 (1973) 977-987. 116 J. Epstein, J. R. Williams and J. B. Little, Rate of DNA repair in progeric and normal human fibroblasts, Biochem. Biophys. Res. Commun., 59 (1974) 850-857.

222 117 J. D. Regan and R. B. Setlow, DNA repair in human progeroid cells, Biochem. Biophys. Res. Commun., 59 (1974) 858-864. 118 T. Kakunaga, Requirement for cell replication in the fixation and expression of the transformed state in mouse cells treated with 4-nitroquinoline-l-oxide, Int. J. Cancer, 14 (1974) 736-742. 119 T. Kakunaga, The role of cell division in the malignant transformation of mouse cells treated with 3-methylcholanthrene, Cancer Res., 35 (1975) 1637-1642. 120 V. M. Maher and J. J. McCormick, Effect of DNA repair on the cytotoxicity and mutagenesis of UV radiation and chemical carcinogens in normal and xeroderma pigmentosum cells, in J. M. Yuhas, R. W. Tennant and J. Regan (eds.), Biology of Radiation Carcinogenesis, Raven Press, New York, 1976, pp. 129-145. 121 V. M. Maher, L. M. Ouelette, R. D. Curren and J. J. McCormick, Frequency of UV-light-induced mutation is higher in xeroderma pigmentosum variant cells than in normal cells, Nature, 261 (1976) 593-595. 122 S. M. D'Ambrosio and R. B. Setlow, Defective and enhanced postreplication repair in classical and variant xeroderma pigmentosum cells treated with N-acetoxy-2-acetylaminofluorene, Cancer Res., 38 (1978) 1147-1153. 123 C. G. Sahasrabuddhe and K. E. van ttolde, The effect of trypsin on nuclease-resistant chromatin fragments, J. Biol. Chem., 249 (1974) 152-156. 124 R. Mandel and G. D. Fasman, Chromatin and nucleosome structure, Nucleic Acids Res., 3 (1976) 1839-1855. 125 G. Moruzzi, B. Baabiroli, M. S. Moruzzi and B. Tadolini, The effect of spermine on transcription of mammalian chromatin by mammalian deoxyribonucleic acid-dependent ribonucleic acid polymerase, Biochem. J., 146 (1975)697- 703. 126 R. C. yon Borstel, D. M. Prescott and F. J. Bollum, Incorporation of nucleotides into nuclei of fixed cells by DNA polymerase,J. CellBiol., 29 (1966) 21-28. 127 S. P. Modak, R. C. von Borstel and F. J. Bollum, Terminal lens cell differentiation, Exp. Cell Res., 56 (1969) 105-112. 128 S. P. Modak and F. J. Bollum, Terminal lens cell differentiation, Exp. Cell Res., 62 (1970) 421432. 129 S. P, Modak and F. J. Bollum, Detection and measurement of single-strand breaks in nuclear DNA in fixed lens sections, Exp. Cell Res., 75 (1972) 307-313. 130 S. P. Modak, W. E. Lever and V. R, R. Uppuluri, Estimation of cell nuclei cut at the surface of tissue sections, Exp. Cell Res., 70 (1972) 465-468. 131 S. P. Modak and H. Traurig, Appearance of strand breaks in nuclear DNA of terminally differentiating vaginal epithelium, Cell Diff., 1 (1972) 351-355. 132 S. P. Modak and G. B, Price, Exogenous DNA polymerase-catalysed incorporation of deoxyribonucleotic monophosphates in nuclei of fixed mouse brain cells, Exp. Cell Res., 65 (1971 ) 289-296. 133 G. B. Price, S, P, Modak and T. Makinodan, Age-associated changes in the DNA of mouse tissue, Science, 171 (1971)917-920. 134 F. J. BoUum, "Primer" in DNA polymerase reactions, Prog. Nucleic Acids Res., 1 (1963) 1-13. 135 M. Yoneda and F. J. Bollum, Deoxynucleotide-polymerizing enzymes of calf thymus gland, J. Biol. Chem., 240 (1965) 3385-3391. 136 J. Piatigorsky, S. S. Rothschild and L. M. Milstone, Differentiation of lens fibers in explanted embryonic chick lens epithelia, Dev, Biol., 34 (1973) 334-345. 137 M. F. Counis, E. Chaudun and Y. Courtois, DNA synthesis and repair in terminally differentiating embryonic lens cells, Dev. Biol., 57 (1977) 47-55. 138 S. P. Modak, Two dimensional electrophoresis of native and denaturated DNA for chromatin digests, Experientia, 34 (1978) 57 (abstract). 139 tt. R. Massie, M. B. Baird and R. J. Nicolosi, Changes in the structure of rat liver DNA in relation to age,Arch. Biochem. Biophys., 153 (1972) 736-741. 140 K. T. Wheeler and J. T. Lett, On the possibility that DNA repair is related to age in non-dividing cells, Proc. Nat. Acad. Sci. U.S.A., 71 (1974) 1862-1865. 141 S. P. Modak and U. Dumaswalla, Native-and-alkaline-agarose gel electrophoresis chromatin subunits derived DNA denatured with acid, heat, DMSO and alkali, (1978) in preparation. 142 P. Karran and M. G. Ormerod, Is the ability to repair damage do DNA related to the proliferative capacity of a cell? The rejoining of X-ray produced strand breaks, Biochim. Biophys. Acta, 299 (1973) 54-64.

223 143 S. Goldstein, The biology of aging, N. Eng. J. Med., 285 (1971) I 120-1129. 144 M. R. Mattern and P. A. Cerutti, Selective excision of gamma ray damaged thymine from the DNA of cultured mammalian cells, Biochim. Biophys. Acta, 395 (1975) 48-55. 145 J. B. Little, Relationships between DNA repair capacity and cellular aging, Gerontology, 22 (1976) 28-55. 146 G. E. Milo and R. W. Hart, Age-related alterations in plasma membrane glycoprotein content and scheduled or unscheduled DNA synthesis, Arch. Biochem. Biophys., 176 (1976) 324-333. 147 G. M. Hahn, D. King and S. J. Yang, Quantitative changes in unscheduled DNA synthesis in rat muscle cells after differentiation, Nature (New BioL), 230 (1971) 242-244. 148 F. E. Stockdale and M. C. O'Neill, Repair DNA synthesis in differentiated embryonic muscle cells, J. CellBiol., 52 (1972) 589-597. 149 A. C. Chan, S. K. C. Ng and I. G. Walter, Reduced DNA repair during differentiation of a myogenic cell line, J. CellBioL, 70 (1976) 685-691. 150 J. Trenton, S. P. Modak and Y. Courtois, Unscheduled DNA synthesis in UV-irradiated embryonic chick lens epithelium and lens fibers, (1978) in preparation. 151 F. E. Stockdale, DNA synthesis in differentiating skeletal muscle cells: initiation by ultraviolet light, Science, 171 (1971) 1145-1147. 152 R. W. Barton and W. K. Yang, Low molecular weight DNA polymerase decreased activity in spleens of old Balb/c mice, Mech. Ageing Dev., 4 (1975) 123-136. 153 S. Linn, M. Kairis and R. Holliday, Decreased fidelity of DNA polymerase activity isolated from aging human fibroblasts, Proc. Nat. Acad. ScL U.S.A., 73 (1976) 2818-2822. 154 M. J. Pythilla and F. G. Sherman, Age-associated studies on thermal stability and template effectiveness of DNA and nucleoproteins from beef thymus, Biochem. Biophys. Res. Commun., 31 (1968) 340-344. 155 H. P. yon Hahn, Structural and functional changes in nucleoprotein during the aging of the cell, Gerontologia, 16 (1970) 116-128. 156 D. I. Kurtz and F. M. Sinex, Age related differences in the association of brain DNA and nuclear protein, Biochim. Biophys. Acta, 145 (1967) 840-842. 157 D. I. Kurtz, A. P. Russel and F. M. Sinex, Multiple peaks in the derivative melting curve ofchromatin from animals of varying age, Mech. Ageing Dev., 3 (1974) 37--49. 158 S. M. Zhelabovskaya and G. D. Berdyshev, Composition, template activity and thermostability of the liver chromatin in rats of various age, Exp. GerontoL, 7 (1972) 313-320. 159 B. T. Hill, Influence of age on chromatin transcription in murine tissues using a heterologous and a homologous RNA polymerase, Gerontology, 22 (I 976) 111-123. 160 H. V. Samis and V. J. Wulff, The template activity of rat liver chromatin, Exp. Gerontol., 4 (1969) 111-117. 161 R. G. Cutler, Transcription of unique and preferated DNA sequences in mouse liver and brain tissues as a function of age, Exp. GerontoL, 10 (1975) 37-60. 162 Zh. A. Medvedev, M. N. Medvedev and L. I. Huschtascha, Age changes of the pattern of F1 histone subfractions in rat liver and spleen chromatin, Gerontology, 23 (1977) 334-341. 163 Zh. A. Medvedev, M. N. Medvedev and L. Robson, Tissue specificity and age changes of the pattern of the H1 group of histones in chromatin from mouse tissues, Gerontology, 24 (1978) 286-292. 164 C. Unger-Ullmann and S. P. Modak, Histones in aging mouse liver, Experientia, 34 (1978) 67 (abstract). 165 D. E. Brash and R. W. Hart, DNA damage and repair in vivo, J. Environ. PathoL ToxicoL, 2 (1978) 79-114. 166 G. A. Martin, Genetic syndromes in man with potential relevance to pathobiology of aging, in D. Harrison (ed.), Proceedings of a Symposium on Genetic Effects on Aging, Academic Press, New York, 1978. 167 R. W. Hart and R. B. Setlow, DNA repair in late-passage human cells, Mech. Ageing Dev., 5 (1976) 67-77. 168 S. P. Modak and C. Unger-UUmann, Control ofgenome integrity in terminally differentiating and aging postmitotic cells, in W. Beermann, W. J. Gehring, J. B. Gurdon, F. C. Katatos, J. Reinest (eds.), Results and Problems in Cell Differentiation, Springer-Verlag, New York, 1979, in press. 169 J. Smith-Sonneborn, DNA repair and longevity assurance in Paramecium tetraurelia, (1978) submitted.