Increased levels of DNA breaks in cerebral cortex of Alzheimer's disease patients.

It has been hypothesized that Alzheimer's disease (AD) is caused by an accumulation of damage in DNA due to defective DNA-repair (21). Attempts to test this hypothesis by determining the activity of DNA-repair systems in nonneuronal cells from AD patients and controls so far provided conflicting results. An alternative approach is the direct comparison of DNA-damage levels in neuronal tissue of AD patients and controls. In the present study we assayed the level of DNA breaks and alkali-labile sites in cerebral cortex tissue samples from AD patients and controls obtained from rapid autopsies. Our data on 11 AD patients and 8 control subjects indicate an at least two-fold higher level of DNA damage in cortex of AD patients as compared to controls.

Decreased DNA repair capacity in familial, but not in sporadic Alzheimer's disease.

Using the alkaline filter elution technique we determined the induction and disappearance of DNA single-strand breaks (SSB) in freshly isolated peripheral blood lymphocytes (PBL) from 43 Alzheimer's disease (AD) patients and 48 normal, healthy age- and sex-matched control subjects following in vitro exposure to N-ethyl-N-nitrosourea (ENU). The mean percentage SSB disappearance in PBL from control subjects at 1 h after ENU treatment was 41.4 +/- 2.9%; this was not significantly different from that found in samples from AD patients which had no (n = 16) or one (n = 12) first-degree relative with dementia (42.5 +/- 8.2% and 43.0 +/- 4.4%, respectively; p greater than 0.75). However, in PBL of 15 AD patients with at least two first-degree relatives with dementia the mean percentage SSB disappearance was 23.6 +/- 5.8%, which was significantly lower than that found in controls (p less than 0.01) or in the other AD patients (p less than 0.02).

Induction and repair of benzo[a]pyrene-DNA adducts in C57BL/6 and BALB/c mice: association with aging and longevity.

In this study, we employed the sensitive 32P-postlabeling assay to assess the influence of age on the formation and disappearance of benzo[a]pyrene (B[a]P) DNA adducts in six organs of two different mouse strains with different life spans, C57BL/6ByJ (C57BL/6) and BALB/cByJ (BALB/c). Following a single, intraperitoneal treatment with 50 mg B[a]P per kg of bodyweight, maximum formation of the major B[a]P-derived adduct, trans-(7R)-N2-[10-(7 beta,8 alpha, 9 alpha-trihydroxy- 7,8,9,10)-tetrahydrobenzo[a]pyrene]-yl-deoxyguanosine (BPDE-N2-dG), appeared to be age- and organ-dependent; minor differences were observed for the same organs between the two mouse strains. The maximum formation of BPDE-N2-dG in the various organs from young and old mice differed by a factor of 2-4 and was two- to eightfold lower in organs from old mice as compared to young mice. The removal of BPDE-N2-dG, up to 7 days after the treatment, was apparently age- and strain-dependent; non-significant differences were observed for organs within strains at each age studied. In young C57BL/6 mice, which have a greater life expectancy than BALB/c, the rate of disappearance of BPDE-N2-dG was significantly higher in liver and heart as compared to young BALB/c. At the older age a decrease in the rate of BPDE-N2-dG disappearance was observed more frequently, and to a relatively greater extent, in organs from C57BL/6 mice as compared to BALB/c mice. These results are discussed in relation to the differences in life spans and the incidence of pathological lesions between the two strains of mice.

Age-dependent accumulation of alkali-labile sites in DNA of post-mitotic but not in that of mitotic rat liver cells.

The amount of spontaneous damage in the DNA of rat liver cells was measured by using the alkaline elution assay. An age-related increase of approximately 700 detectable alkali-labile sites (80%) was found for rat parenchymal liver cells; cells from 6-month-old rats contained approximately 900 alkali-labile sites per cell while cells from 36-month-old rats contained approximately 1600 alkali-labile sites. In contrast to the situation with the postmitotic parenchymal liver cells, no age-related increase in the number of alkali-labile sites was found for the non-parenchymal liver cell fraction, which has a higher mitotic activity. These results support the hypothesis that aging takes place predominantly in postmitotic cells.

Transgenic mouse models for studying mutations in vivo: applications in aging research.

To study mutation accumulation in the DNA of somatic cells and tissues during aging in vivo, a transgenic mouse model has been constructed. The model harbors plasmid vectors, containing the lacZ reporter gene, integrated head to tail at various chromosomal locations. Procedures have been worked out to efficiently recover the plasmids into E. coli host cells. A positive selection system, permitting only E. coli cells with a lacZ mutated plasmid to grow, allows for the accurate determination of mutation frequencies as the ratio of mutant colonies versus the total number of transformants, i.e., the total number of plasmid copies recovered. Results obtained from a life span study of plasmid mice with vector clusters on chromosome 3 and 4 indicated age-related mutation accumulation in the liver, but not in the brain. Comparison of the mutational spectra revealed a significantly larger proportion of large size-change mutations in liver than in brain.

Transgenic mouse models for studying mutations in vivo: applications in aging research.

To study mutation accumulation in the DNA of somatic cells and tissues during aging in vivo, a transgenic mouse model has been constructed. The model harbors plasmid vectors, containing the lacZ reporter gene, integrated head to tail at various chromosomal locations. Procedures have been worked out to efficiently recovery the plasmids into E. coli host cells. A positive selection system, permitting only E. coli cells with a lacZ mutated plasmid to grow, allows for the accurate determination of mutation frequencies as the ratio of mutant colonies versus the total number of transformants, i.e., the total number of plasmid copies recovered. Results obtained from a life span study of plasmid mice with vector clusters on chromosome 3 and 4 indicated age-related mutation accumulation in the liver, but not in the brain. Comparison of the mutational spectra revealed a significantly larger proportion of large size-change mutations in liver than in brain.

Carnitine palmitoyl transferase-I activity in the aging mouse heart.

We investigated the influence of age on carnitine palmitoyl transferase-I (CPT-I, EC 2.3.1.21) activity in the mouse heart. There was an age-associated decrease in CPT-I activity from 2 to 26 months (P = 0.006). We studied the effect of oxygen-derived radicals on CPT-I activity. Mitochondria from 2-month-old mouse hearts exposed to different concentrations of hydrogen peroxide (H2O2) showed a dose-related decrease in CPT-I activity (P < 0.002). To determine the possible reversibility of the age change in CPT-I activity, we studied the effect of oral administration of propionyl-L-carnitine (PLC). Oral pretreatment of middle-aged (18-month-old) mice with PLC resulted in a 37% increase of basal CPT-I activity (P < 0.05) compared to age-matched untreated animals, and restored it to a level similar to that of 2-month-old mice. Pretreatment of senescent (26-month-old) mice with PLC, however, showed no significant change in basal CPT-I activity. It is possible that the age-related decrease in CPT-I activity may result from an in vivo accumulation of oxygen-derived radical damage. It appears that the age change in CPT-I activity in 18- but not in the 26-month-old mice is reversible with PLC.

Treatment of lacZ plasmid-based transgenic mice with benzo[a]pyrene: measurement of DNA adduct levels, mutant frequencies, and mutant spectra.

The evaluation of the relationship between the dose to DNA of a genotoxic carcinogen and in vivo somatic cell mutagenesis may provide important information on the mechanisms leading to induced mutational events. The induction of DNA adducts and DNA mutations in different tissues of lacZ plasmid-based transgenic mice has been investigated following a single intraperitoneal administration of the polycyclic aromatic hydrocarbon benzo[a]pyrene (B[a]P). DNA adducts were measured by (32)P-postlabeling at times between 1 and 28 days following injection and DNA mutations in the lacZ reporter gene were quantified using a highly efficient immunomagnetic purification of the lacZ-containing pUR288 plasmid, followed by electrotransformation of circularized plasmids into a host-restriction negative E. coli C (DeltalacZ, galE (-)) host. The major DNA adduct formed by B[a]P, N (2)-(10beta-[7beta,8alpha, 9alpha-trihydroxy-7,8,9, 10-tetrahydrobenzo(a)pyrene]yl-deoxyguanosine, reached maximum levels between 5 and 7 days followed by a gradual decrease. The induced mutant frequency reached a maximum between 7 and 14 days, after the DNA adduct level in a particular tissue had reached its apparent maximum between 5 and 7 days. The mutant spectrum, defined as the percentage of no-change and size-change mutations in a particular tissue, changed from predominantly size-change mutations in untreated tissues to predominantly no-change mutations in tissues displaying the highest B[a]P-induced mutant frequencies. Contrary to the assumption that there are no factors that can eliminate mutations once they exist, it was observed that mutant frequencies in the organs studied declined to background levels, which was accompanied by a change in the mutant spectrum from predominantly no-change mutations to predominantly size-change mutations.

High sensitivity for color mutants in lacZ plasmid-based transgenic mice, as detected by positive selection.

Transgenic mice carrying bacteriophage lambda-based vectors harboring a lacZ or lacI reporter gene have been used in recent years for the detection, quantification, and characterization of spontaneous and induced gene mutations in vivo. The usefulness of these models is basically determined by their ability to detect significant (organ-specific) exposure-related increases in mutant frequencies. This, in its turn, is dependent on the models' ability to detect all possible types of mutations, including mutations that partially inactivate the gene product, as well as the absence of a significant proportion of E. coli-derived mutations. The newly developed lacZ plasmid-based transgenic mouse mutation assay is sensitive to a broad range of DNA mutations, including large size changes. Here, we describe the plasmid model's sensitivity for detecting color mutants, i.e., cells containing a partially inactivated lacZ gene, as detected by a positive selection system using phenyl-beta-galactoside. The proportion of color mutants in brain and kidney was 30-40%. This was significantly higher than the 10-20% color mutants observed in lung, spleen, and liver. It is demonstrated that the color mutants detected were neither the result of a mixture of both mutant and nonmutant cells within the original mutant colony-forming unit nor due to the presence of mutant and nonmutant lacZ genes within individual E. coli host cells. Sequence analysis of 13 different color mutants revealed single basepair substitutions in the lacZ gene of each mutant. The high tolerance of the positive selection system for lacZ color mutants in the plasmid-based transgenic mouse model greatly contributes to the sensitivity of this model, now ranging from large size-change mutations that completely inactivate the lacZ gene to basepair substitutions that partly inactivate the lacZ gene.

Sources of variability in mutant frequency determinations in different organs of lacZ plasmid-based transgenic mice: experimental features and statistical analysis.

Sources of variability that may affect the measurement of mutant frequencies in lacZ plasmid-based transgenic mice, including rescue-to-rescue, animal-to-animal (within organ), and organ-to-organ, were studied in the context of some of the experimental features that distinguish this model from other currently used systems. Statistical analysis of repeated determinations of DNA from kidney, spleen, liver, lung, and brain indicated that 350,000 colony-forming units from each of three animals should be analyzed per treatment group in order to detect a 50% difference in mutant frequencies. Consideration is given to some of the experimental features of the transgenic assay system, including its positive selection system, its rescue efficiency, the ability to detect large deletions in the lacZ target genes, the contribution of Escherichia coli-derived mutations to the spontaneous mutant frequencies observed in vivo, and cost effectiveness of mutant frequency determinations.

In vivo transgenic mutation assays.

Transgenic rodent gene mutation models provide quick and statistically reliable assays for mutations in the DNA from any tissue. For regulatory applications, assays should be based on neutral genes, be generally available in several laboratories, and be readily transferable. Five or fewer repeated treatments are inadequate to conclude that a compound is negative but more than 90 daily treatments may risk complications. A sampling time of 35 days is suitable for most tissues and chemicals, while shorter sampling times might be appropriate for highly proliferative tissues. For phage-based assays, 5 to 10 animals per group should be analyzed, assuming a spontaneous mutant frequency (MF) of approximately 3 x 10(-5) mutants/locus and 125,000-300,000 plaque or colony forming units (PFU or CFU) per tissue. Data should be generated for two dose groups but three should be treated, at the maximum tolerated dose (MTD), two-thirds the MTD, and one-third the MTD. Concurrent positive control animals are only necessary during validation, but positive control DNA must be included in each plating. Tissues should be processed and analyzed in a block design and the total number of PFUs or CFUs and the MF for each tissue and animal reported. Sequencing data would not normally be required but might provide useful additional information in specific circumstances. Statistical tests used should consider the animal as the experimental unit. Nonparametric statistical tests are recommended. A positive result is a statistically significant dose-response and/or statistically significant increase in any dose group compared to concurrent negative controls using an appropriate statistical model. A negative result is statistically nonsignificant with all mean MF within two standard deviations of the control.

Single-strand break disappearance in quiescent and phytohaemagglutinin-stimulated human peripheral blood lymphocytes exposed to a single low dose of gamma-radiation.

Quiescent and phytohaemagglutinin (PHA)-stimulated human peripheral blood lymphocytes (PBL) were irradiated with 4 Gy of gamma-rays and assayed using the alkaline filter elution technique to determine (1) the rate of removal of single-strand breaks (ssb) and (2) the occurrence of excision repair events as indicated by the accumulation of ssb in the presence of the excision repair inhibitor 1-beta-D-arabinofuranosylcytosine (araC). The percentage of ssb disappearance, in the absence of araC, at 5 min after irradiation was significantly higher in PHA-stimulated PBL than in quiescent PBL [40.4 +/- 8.4% (mean +/- SD) and 71.3 +/- 6.8% in quiescent and PHA-stimulated PBL, respectively]. In the presence of araC, both quiescent and PHA-stimulated PBL rapidly accumulated araC-associated ssb, indicating the inhibition of early (base excision) repair processes acting on alkali-stable base damages. Results with PBL from two different donors indicated a significantly higher rate of accumulation of araC-associated ssb in PHA-stimulated PBL than in quiescent cells. In PBL from a third donor no such difference in the rate of accumulation of araC sites was observed. After 1 h repair incubation, the same number of araC-associated ssb was found in the two different cell populations from all three donors.

Age-related induction and disappearance of carcinogen-DNA-adducts in livers of rats exposed to low levels of 2-acetylaminofluorene.

It was investigated whether in vivo aging of rat liver is associated with changes in the induction and rate of disappearance of DNA damage. For this purpose 6- and 36-month-old rats were intraperitoneally injected with a single, low dose (5 mg/kg body wt.) of the model liver carcinogen 2-acetylaminofluorene (AAF). Using the 32P-postlabeling assay we found that N-(deoxyguanosin-8-yl)-2-aminofluorene (dG-C8-AF) was the major DNA-adduct formed. The minor adduct N-(deoxyguanosin-8-yl)-2-acetylaminofluorene (dG-C8-AAF) could only be detected after doses of 20 mg/kg or more. Quantitation of adduct levels at various time points after treatment indicated a rapid induction of AF-adducts, which were already present at 6 h after treatment. The subsequent loss of AF-adducts was relatively slow, as was indicated by the presence of a substantial amount of AF-adducts as late as 21 days after treatment. Slight age-related differences in the pattern of induction and disappearance of AF-adducts and a somewhat higher level of persisting lesions in old than in young rats were observed.

Induction and disappearance of DNA single-strand breaks in human B and T lymphocytes after exposure to ethylnitrosourea.

Using the alkaline filter elution technique we monitored the induction and disappearance of DNA single-strand breaks (SSB) in 3 different human lymphocyte populations: (1) freshly isolated peripheral blood lymphocytes (PBL); (2) B and T cell-enriched lymphocyte fractions; and (3) actively proliferating T cells, after exposure to ethylnitrosourea (ENU). Between these different lymphocyte populations no significant differences were observed in the number of SSB induced by a 20-min treatment with 0.5 mM ENU. SSB disappearance was observed in PBL of some but not all individuals, confirming our earlier results (Boerrigter et al., 1990a). Determinations on B and T cell-enriched lymphocyte populations indicated that ENU-induced SSB were removed only in T lymphocytes; no significant amount of SSB disappearance was observed in B lymphocytes. In contrast, no differences in SSB repair between B and T lymphocytes were found after gamma-irradiation. Induction and disappearance of ENU-induced SSB were found not to be dependent on the proliferative status of T lymphocytes; no differences were observed between quiescent PBL or T lymphocytes and actively proliferating T cells from the same donor, with respect to either the rate or the total amount of ENU-induced SSB disappearance.

Spontaneous DNA breaks in the rat brain during development and aging.

The level of spontaneous DNA breaks in nuclei isolated from the cerebral cortex of rat embryos at 12, 15 and 19 days of gestation, and from cerebral cortex and cerebellum of 24-day-, 6-month- and 36-month-old rats was measured by alkaline elution. A constant low level of DNA breaks was found in brain DNA during development from an embryo at day 12 of gestation to a 24-day-old rat. During aging the level of DNA breaks remained at the same low level, as shown by comparing nuclei from the cerebral cortex and cerebellum of 6- and 36-month-old animals. By contrast, an almost 2-fold increase in the level of DNA breaks was observed in rat liver nuclei between 6 and 36 months of age, confirming our earlier findings on isolated liver cells. Although there were no changes in the level of DNA breaks in rat brain during development or during aging, breaks accumulated rapidly post mortem. The rate of this process was not age-dependent. Our data suggest that the level of spontaneous DNA breaks in the brain is not likely to be of fundamental importance in the complex cellular alterations associated with brain development and aging.

Use of transgenic mouse models for studying somatic mutations in aging.

Theories on the causes of aging, based on the accumulation of somatic mutations in tissues of an organism, were formulated decades ago, but remain insufficiently tested. Transgenic animals, equipped with integrated bacterial reporter genes that can be efficiently rescued from total genomic DNA of all tissues and organs, represent ideal tools for investigating the types and frequencies of spontaneous mutants accumulating during aging. The first of such systems, based on the transgenic integration of bacteriophage lambda shuttle vectors that contain the bacterial lacZ gene as mutational target, was constructed in our laboratory and is now routinely used. Results obtained with this and the related LacI system that are relevant for the somatic mutation theory of aging will be discussed. One conclusion is that, due to the nature of the transgene, lambda-based systems have the disadvantage that deletion type mutations are underrepresented in comparison to point mutations. To overcome those limitations, we constructed a new transgenic mouse model carrying a pUR288 plasmid shuttle vector with the lacZ reporter gene. Some preliminary data obtained with this model serve to illustrate its potential use to extensively test the somatic mutation theory of aging.

Investigation of the radioadaptive response in brain and liver of pUR288 lacZ transgenic mice.

The radioadaptive response, where a small priming dose of ionizing radiation can lessen the effects of subsequent exposure to a higher radiation challenge dose, was investigated in brain and liver within transgenic mice. Although it is well characterized in models in vitro, current radioadaptive response research has focused on particular cell types (i.e., lymphocytes) and does not provide comparative data for responses of multiple tissues within an organism. Transgenic animals are useful for such comparisons, because the transgene is integrated into all cells in the body. The pUR288 lacZ plasmid-based transgenic mouse model utilizes a plasmid vector allowing highly efficient recovery of mutational targets, including large size-change mutations that result from radiation exposure. Female C57BI/6 pUR288 lacZ mice were exposed to priming doses of 0.075- to 0.375-Gy x-rays over a 3-d period. After 3 wk, they received an acute challenge dose of 2.5-Gy x-rays. Spontaneous mutant frequencies in lacZ were significantly higher in liver than in brain (6.62 x 10(-5) vs. 3.51 x 10(-5)). In the absence of a priming dose, the 2.5-Gy challenge doubled the mutant frequency of both liver and brain (13.38 x 10(-5), and 7.63 x 10(-5) respectively). Priming doses of 0.15, 0.225, and 0.375 Gy significantly reduced (by 40%) the mutagenic effects of the 2.5-Gy challenge in brain. Restriction enzyme analysis of rescued mutant plasmids revealed a decrease in large size-change mutations at the three priming doses in brain. This study demonstrates the utility of this model for the investigation of radiological processes of large size-change mutations, as well as showing a radioadaptive response in brain, but not liver, of mice in vivo.

DNA strand break metabolism in human lymphocytes: a reevaluation.

Recent findings concerning the presumed existence of single-strand breaks (SSB) in quiescent human peripheral blood lymphocytes (PBL) are discussed in relation to the role of poly(ADP-ribosyl)ation in DNA strand break metabolism. It is argued that the activation of poly(ADP-ribose)polymerase (ADPRP) by a DNA-damaging agent is not indicative of an obligatory role of poly(ADP-ribosyl)ation in DNA repair. From this it follows that SSB induced by different strand-breaking agents might be removed by either ADPRP-dependent or ADPRP-independent DNA repair pathways.

Studies on DNA repair defects in degenerative brain disease.

Using the alkaline filter elution technique we determined the induction and disappearance of single-strand breaks (SSB) in freshly isolated peripheral blood lymphocytes (PBL) from 43 Alzheimer disease (AD) patients and from 48 healthy, age- and sex-matched control subjects following in vitro exposure to N-ethyl-N-nitrosourea (ENU), methyl methanesulphonate (MMS), or gamma (gamma)-radiation. No differences in SSB disappearance between AD patients and controls were observed after treatment of PBL with MMS or gamma-rays. After treatment with ENU, however, the amount of SSB disappearance was significantly lower in PBL from familial, but not in PBL from sporadic AD patients. ENU repair in PBL from neurological controls was comparable to that found in normal age-matched controls, indicating that the lower amount of ENU repair in familial AD patients is not a consequence of neuronal degeneration. These tentative findings are discussed in relation to the aetiology of AD.

DNA repair and Alzheimer's disease.

The accumulation of nonrepaired or misrepaired DNA lesions, induced by exogenous and/or endogenous DNA-damaging agents, has been postulated to be one of the major and fundamental causes of aging. The accumulation of DNA damage might lead to changes in the expression of genes important for cell survival and, as such, could be a relevant risk factor in the etiology of various age-related diseases, including neurodegenerative diseases. This review deals specifically with Alzheimer's disease (AD) and the proposed role of defective DNA repair in the etiology of this disease. The evidence that a genetic defect in some aspect of base excision repair is present in cells from AD patients is discussed. Based on the available data, it seems justified to conclude that in at least the familiar form of the disease such a DNA repair defect is present in fibroblast and lymphoid cells. The absence of sufficiently efficient DNA repair mechanisms, involved in the removal of small base damages in the brain, could result in the accumulation of misrepaired or nonrepaired DNA damage and might ultimately lead to the neuronal degeneration as observed in AD patients. In this respect, the existence of inherently imperfect DNA repair processes in longer-lived species, such as mammals, might be regarded as a trade-off between reproductive success and cellular maintenance and repair, which would lead to a high level of DNA damage and mutation at old age.