Mitochondrial genome instability in cancer.

The relevant role of mitochondrial mutations in cancer is the most frequent conclusion found in most early publications on the subject. However, it is now clear that this assumption was in many cases based on circumstantial or even flawed evidence. Presently, we know that normal mitochondria structure and functions depend on the concerted interaction between mitochondrial (mt)-genes and different groups of nuclear genes. Thus, somatic mutations of mt- or nuclear genes controlling mitochondrial physiology would influence the cancer transformation process through a disruption of nuclear<-->mitochondrial gene interactions. In this regard, somatic mt-mutations influencing carcinogenesis have been detected in preneoplastic lesions. Furthermore, an abnormal respiration process with the subsequent increase in reactive oxygen species production seems to be one of the basic mechanisms favoring oncogenesis. Many mt-genes exhibit inherited polymorphisms associated with their mitochondrial phylogenetic history. In this report we shall summarize data showing that some of these ethnic mt- mutations may increase or alternatively decrease the susceptibility to various forms of malignancy. The interference of mt-mutations with anticancer therapies and the use of body fluids for the analysis of mt-mutations to obtain tumor samples avoiding invasive techniques are two promising fields to be further investigated.

Allele and genotype frequencies of metabolic genes in Native Americans from Argentina and Paraguay.

Interethnic differences in the allele frequencies of CYP2D6, NAT2, GSTM1 and GSTT1 deletions have been documented for Caucasians, Asians, and Africans population. On the other hand, data on Amerindians are scanty and limited to a few populations from southern areas of South America. In this report we analyze the frequencies of 11 allele variants of CYP2D6 and 4 allele variants of NAT2 genes, and the frequency of GSTM1 and GSTT1 homozygous deleted genotypes in a sample of 90 donors representing 8 Native American populations from Argentina and Paraguay, identified as Amerindians on the basis of their geographic location, genealogical data, mitochondrial- and Y-chromosome DNA markers. For CYP2D6, 88.6% of the total allele frequency corresponded to *1, *2, *4 and *10 variants. Average frequencies for NAT2 *4, *5, *6 and *7 alleles were 51.2%, 25%, 6.1%, and 20.1%, respectively. GSTM1 deletion ranged from 20% to 66%, while GSTT1 deletion was present in four populations in less than 50%. We assume that CYP2D6 *2, *4, *10, *14; NAT2 *5, *7 alleles and GSTM1 and GSTT1 *0/*0 genotypes are founder variants brought to America by the first Asian settlers.

Ethnic variation in the prevalence of AZF deletions in testicular cancer.

Seventy-six percent of testicular cancers of origin in Finns have been reported to exhibit AZF deletions. We analyze here 40 testicular tumor cases from Norway and Argentina and we found that AZF deletions occur also in non-Finnish cases but at significantly lower frequency (25%) than in Finland testicular tumor cases. This frequency difference can be attributed to the condition of genetic isolate of the Finnish population and the subsequent prevalence in this ethnic group of genetic factors involved in the origin of AZF deletions associated with malignancies. The finding of a correlation between AZF deletions and a given Y haplogroup would indicate the existence of Y lineages carrying AZF deletion-enhancing gene or genes. This possibility was explored using a set of Y-DNA-markers allowing the identification of Y lineages occurring at high frequency in Finns. We characterized the Y haplogroups in 32 normal Finn males (control group) and 17 cases of testicular cancer in Finns with and without AZF deletions. We found no association between Y lineages and AZF microdeletions, nor between AZF microdeletions and a Y microdeletion (DYS7C) exhibiting high prevalence (>50%) in Finns. The lack of correlation between AZF deletions and Y haplogroups indicates that the origin of these deletions is independent from the Y chromosome genetic background. No AZF deletions were found in familial cases of testicular tumors; hence, hereditary factors inducing the appearance of testicular malignancies in certain genealogies apparently do not increase the susceptibility to AZF deficiencies. AZF deletions are de novo events occurring in prezygotic or in post-zygotic stages. We propose that most AZF deletions associated with testicular tumors are due to post-zygotic Y microdeletions, while most cases of deletions associated with infertility are due to deletions occurring in the germ cell line of proband fathers.

Akodon sex reversed females: the never ending story.

The existence of fertile A. azarae females with a chromosome sex pair indistinguishable from that of males was reported more than 35 years ago. These heterogametic females were initially thought to occur due to an extreme process of dosage compensation in which X inactivation was restricted to Xp and complemented by a deletion of Xq (Xx females). Later on, a C-banding analysis of A. mollis variant females showed that these specimens were in fact XY* sex reversed and not Xx females. The finding of positive testing for Zfy and Sry multiple-copy genes in Akodon males and heterogametic females confirmed the XY* assumption. At the present time, XY* sex reversed females have been found to exist in nine Akodon species. Akodon heterogametic females produce X and Y* oocytes, which upon sperm fertilization give rise to viable XX (female), XY* (female), and XY (male) embryos, and to non-viable Y*Y zygotes. Heterozygous females exhibit a better reproductive performance than XX females in order to compensate the Y*Y zygote wastage. XY* sex reversed females are assumed to occur due to a deficient Sry expression resulting in the development of ovaries instead of testes. Moreover, the appearance of Y* elements is a highly recurrent event. It is proposed that homozygosity for an autosomal or pseudoautosomal recessive mutation (s-) inhibits Sry expression giving rise to XY* embryos with ovary development. Location of the Y* chromosome in the female germ cell lineage produces an ovary-specific imprinting of the Sry* gene maintaining its defective expression through generations independently from the presence or absence of s- homozygosity. By escaping the ovary-specific methylation some Y* chromosomes turn back to normal Ys producing Y oocytes capable of generating normal male embryos when fertilized by an X sperm. Fluctuations in the rate of variant females in field populations and in laboratory colonies of Akodon depend on the balance between the appearance of new variant females (s-/s-, XY* specimens) and the extinction of sex reversed specimens due to imprinting escape.

Correlation between molecular and conventional genealogies in Aicuña: a rural population from Northwestern Argentina.

Aicuña is a village in the northwest of Argentina, located about 300 km south of La Rioja city, in the province of La Rioja. The population of Aicuña derives from a founder couple established in the uninhabited Aicuña valley in the early years of the 17th century. Due to land ownership litigation, the descendants maintained a well-documented genealogy that extends for 12 generations, comprising more than 8,000 individuals. From the historical pedigree of Aicuña, we selected 14 males with direct patrilineal descent from the 2 most ancient male founders, and 23 donors (9 females and 14 males) with direct matrilineal descent from the most ancient female founder. All 3 founders lived in the 17th century. We collected DNA from buccal swabs and characterized the mitochondrial DNA (mtDNA) and Y haplotypes using 14 Y-specific markers, 11 mtDNA polymorphic markers and sequencing of the mt hypervariable regions 1 and 2. We found four different Y haplotypes: Y1 and Y2 haplotypes of European origin corresponding to the founder ancestors Francisco Páez de Espinoza and Apolinario Ormeño, which were shared by 6 and 3 donors, respectively. Three males selected as Ormeño patrilineal descendants showed a different Y haplotype (Y3), probably originated by erroneous paternity registration due to illegitimacy. The remaining case (haplotype Y4), also assumed to belong to the Ormeño lineage, was probably also due to an erroneously registered paternity. Twenty-two donors showed an association of mtDNA markers corresponding to the Amerindian haplotype A2. The founder of this matrilineage could be traced back for more than 14 generations. The haplotype B of one remaining female did not correspond with the historical pedigree and could be due to an error in the genealogy registration. Our results showed an 85% agreement between conventional and molecular genealogies, with mtDNA markers being Amerindian, and Y markers being European. The methodology used in this report is a tool which could potentially be employed as a precedent for land ownership by Aicuña villagers and Amerindian populations.

Mitochondrial genome instability in human cancers.

Malfunction of mismatch repair (MMR) genes produces nuclear genome instability (NGI) and plays an important role in the origin of some hereditary and sporadic human cancers. The appearance of non-inherited microsatellite alleles in tumor cells (microsatellite instability, MSI) is one of the expressions of NGI. We present here data showing mitochondrial genome instability (mtGI) in most of the human cancers analyzed so far. The mtDNA markers used were point mutations, length-tract instability of mono- or dinucleotide repeats, mono- or dinucleotide insertions or deletions, and long deletions. Comparison of normal and tumoral tissues from the same individual reveals that mt-mutations may show as homoplasmic (all tumor cells have the same variant haplotype) or as heteroplasmic (tumor cells are a mosaic of inherited and acquired variant haplotypes). Breast, colorectal, gastric and kidney cancers exhibit mtGI with a pattern of mt-mutations specific for each tumor. No correlation between NGI and mtGI was found in breast, colorectal or kidney cancers, while a positive correlation was found in gastric cancer. Conversely, germ cell testicular cancers lack mtGI. Damage by reactive oxygen species (ROS), slipped-strand mispairing (SSM) and deficient repair are the causes explaining the appearance of mtGI. The replication and repair of mtDNA are controlled by nuclear genes. So far, there is no clear evidence linking MMR gene malfunction with mtGI. Polymerase gamma (POLgamma) carries out the mtDNA synthesis. Since this process is error-prone due to a deficiency in the proofreading activity of POLgamma, this enzyme has been assumed to be involved in the origin of mt-mutations. Somatic cells have hundreds to thousands of mtDNA molecules with a very high rate of spontaneous mutations. Accordingly, most somatic cells probably have a low frequency of randomly mutated mtDNA molecules. Most cancers are of monoclonal origin. Hence, to explain the appearance of mtGI in tumors we have to explain why a given variant mt-haplotype expands and replaces part of (heteroplasmy) or all (homoplasmy) wild mt-haplotypes in cancer cells. Selective and/or replicative advantage of some mutations combined with a severe bottleneck during the mitochondrial segregation accompanying mitosis are the mechanisms probably involved in the origin of mtGI.

Nuclear and mitochondrial genome instability in human breast cancer.

We analyzed 40 pairs of breast normal/cancer tissues for the presence of mitochondrial (mt) genome instability and nuclear MSI in tumor cells. As mt, markers we used a (CA)n mt microsatellite (MS) starting at the 514-bp position of the D loop region and 4 informative MnlI sites located between the 16,108- and 16,420-bp positions of the D loop region. Nuclear microsatallite instability (MSI) was tested with 8 (CA)n MS, syntenic for the 13q chromosome arm. Moreover, we tested the spontaneous frequency of mtMSI and mt-MnlI mutations in 459 mother/descendant events. Mutations of mt-MnlI sites were found in 19 of 40 (47.5%) breast tumors, representing a 216-fold increase over the spontaneous rate in the female germline. Instability of the mtMS occurred in 17 of 40 (42.5%) breast cancers, which implies a 16-fold increase over the rate of spontaneous mutations. Nuclear MSI was found in 20 of 40 (50%) cases. In 15 of these cases the MSI was restricted to one locus, whereas in 5 instances the change of alleles was detected in 2 or 3 loci. Analysis of the correlation between mt and nuclear mutations showed no significant associations, suggesting that different systems are responsible for mt and nuclear genome instability in tumor cells. We propose that the two main mechanisms producing mtRFLP and mtMSI are damage by free radicals and error repair by the polymerase gamma, the first mechanism being a major cause of MnlI mutations and a secondary cause of mtMSI.

Analysis of telomeric repeats and telomerase activity in human colon carcinoma cells with gene amplification.

COLO320DM and COLO320HSR are cell lines derived from a human malignant neuroendocrine colon carcinoma. Both lines have a 30-40-fold amplification of a large DNA domain containing the MYC oncogene. By using fluorescence in situ hybridization techniques with a MYC probe, we could demonstrate that MYC amplicons are contained in a large marker chromosome in COLO320HSR cells, in double minutes (dmin) of COLO320DM cells, and in the interstitial regions of 3-4 additional chromosomes in both cell lines. Amplicons in homogeneous staining regions (HSRs) comprise normal MYC genes, while dmin chromosomes contain PVT/MYC chimeras. Although both cell lines showed similar levels of telomerase activity, the telomere length and telomere distribution in chromosomal termini were considerably lower in COLO320DM than in COLO320HSR cells. This indicates that the average telomere length in cancer cells is regulated no only by the rates of telomerase activity but also by some other non-enzymatic mechanisms.

Origin of YAP+ lineages of the human Y-chromosome.

We screened a total of 841 Y-chromosomes representing 36 human populations of wide geographical distribution for the presence of a Y-specific Alu insert (YAP+ chromosomes). The Alu element was found in 77 cases. We tested 5 biallelic and 8 polyallelic markers in 70 out of the 77 YAP+ chromosomes. We could identify the existence of a hierarchical and chronological structuring of ancestral and derived YAP+ lineages, giving rise to 4 haplogroups, 14 subhaplogroups and 60 haplotypes. Moreover, we propose a monophyletic origin for each one of the YAP+ lineages. Out-of-Africa and out-of-Asia models have been suggested to explain the origin and evolution of ancestral and derived YAP+ elements. We analyze the evidence supporting these two hypotheses, and we conclude that the information available does not allow one to decide between the out-of-Asia or out-of-Africa models.

Characterization of ancestral and derived Y-chromosome haplotypes of New World native populations.

We analyze the allelic polymorphisms in seven Y-specific microsatellite loci and a Y-specific alphoid system with 27 variants (alphah I-XXVII), in a total of 89 Y chromosomes carrying the DYS199T allele and belonging to populations representing Amerindian and Na-Dene linguistic groups. Since there are no indications of recurrence for the DYS199C-->T transition, it is assumed that all DYS199T haplotypes derive from a single individual in whom the C-->T mutation occurred for the first time. We identified both the ancestral founder haplotype, 0A, of the DYS199T lineage and seven derived haplogroups diverging from the ancestral one by one to seven mutational steps. The 0A haplotype (5.7% of Native American chromosomes) had the following constitution: DYS199T, alphah II, DYS19/13, DYS389a/10, DYS389b/27, DYS390/24, DYS391/10, DYS392/14, and DYS393/13 (microsatellite alleles are indicated as number of repeats). We analyzed the Y-specific microsatellite mutation rate in 1,743 father-son transmissions, and we pooled our data with data in the literature, to obtain an average mutation rate of.0012. We estimated that the 0A haplotype has an average age of 22,770 years (minimum 13,500 years, maximum 58,700 years). Since the DYS199T allele is found with high frequency in Native American chromosomes, we propose that 0A is one of the most prevalent founder paternal lineages of New World aborigines.

Paternal directional mating in two Amerindian subpopulations located at different altitudes in northwestern Argentina.

We used mitochondrial DNA (mtDNA) and Y-chromosome DNA polymorphisms to analyze the ethnic origin of maternal and paternal lineages in two Amerindian subpopulations from northwestern Argentina. One of the subpopulations was from San Salvador de Jujuy, located 1200 m above sea level. The second subpopulation inhabits the Quebrada de Humahuaca area at altitudes ranging from 2500 to 3500 m. Both subpopulations have the same ethnic background. All mtDNA haplotypes were identified as Amerindian with a frequency of 64.6% of the B form (9-bp deletion in mtDNA region V). Because all Central Andean Amerindian populations studied so far exhibit high frequencies of the B haplotype, we propose that they probably are derived from a common ancestral population that inhabited the Central Andes 6000-8000 years B.P. The presence of paternal directional mating (asymmetric contribution of one parental lineage to interethnic gene mixtures) was demonstrated by the finding of an average introgression of 40.5% Spanish Y chromosomes into our Amerindian sample. This introgression was more evident at low altitude than at high altitude, with frequencies of 64.3% in San Salvador de Jujuy (low altitude) and 27.6% in Quebrada de Humahuaca (high altitude) (p < 0.05). The San Salvador de Jujuy subpopulation also showed a significantly higher Y-chromosome gene variability than the Quebrada de Humahuaca subpopulation. These findings are in good agreement with historical reports indicating that the colonization of South America was undertaken by men who usually practiced polygamous unions with Amerindian women and that San Salvador de Jujuy was the main northwestern Argentinian region of European to Amerindian gene admixture. We found 16.7% of cases with Spanish Y chromosomes and Amerindian family names, and the same percentage with Amerindian Y chromosomes and Hispanic names. The former group probably is the result of unions between Hispanic men, who transmitted the Y chromosome, and Amerindian women, who transmitted the family name to the progeny. The latter group likely illustrates the practice of changing names from Amerindian to Hispanic during the baptism of native Americans in colonial times.

Effects of antioxidants on streptozotocin-induced clastogenesis in mammalian and insect cells.

The effects of the antioxidant enzymes superoxide dismutase (SOD), catalase (CAT) and the hydroxyl radical scavenger mannitol (MAN) on the clastogenesis induced by STZ in Chinese hamster ovary (CHO) and mosquito cells were investigated. The addition of liposome-entrapped SOD, CAT and MAN to both cell lines caused a significant decrease in the yield of STZ-induced chromosome aberrations (p<0.05). However, the inhibitory effect was more evident in CHO (40.6-67.5%) than in mosquito (15.2-53.6%) cells. These findings indicate that the chromosome damage induced by STZ can be partially inhibited through the incorporation of antioxidant compounds into the cells and suggest that free radicals are involved in the clastogenesis by STZ.

Chromosomal response of insect and mammalian cells to streptonigrin: a comparative study.

We assessed the chromosomal response of insect (mosquito, Aedes albopictus) and mammalian (Chinese hamster ovary, CHO) cells to streptonigrin (SN). Both types of cells were pulse-treated for 20 min with increasing doses of SN and the frequency of chromosome aberrations and sister chromatid exchanges (SCEs) for each SN dose was determined. Our results show that the SN doses inducing remarkable chromosome damage (expressed as frequency of aberrations per cell and per chromosome) in CHO cells fail to produce a significant increase of aberrations in mosquito chromosomes. Moreover, CHO cells exhibited a dose-dependent increase in SCEs which was not observed in mosquito cells. Our results show that while mammalian cells are very sensitive, insect cells are very resistant to SN at the chromosome level. It is possible that variations in the chromatin fibril structure and in the intracellular antioxidant pool may be responsible for the differential response of insect and mammalian chromosomes to SN.

Characterization of mitochondrial DNA and Y-chromosome haplotypes in a Uruguayan population of African ancestry.

We used a set of informative mtDNA and Y-chromosome-specific markers to determine the origin of maternal and paternal lineages in a sample of 41 Uruguayan black individuals. We found that 20 maternal lineages were African, 13 were Amerindian, and 5 were Caucasian. In three individuals we were unable to determine the ethnic origin of the mtDNA lineages. Of the 22 males analyzed we found 4 Y chromosomes of African origin, 5 of Caucasian origin, and 13 of undetermined ancestry. Our results suggest that mtDNA and Y-chromosome-specific DNA variants may be a useful tool in determining the level of mtDNA and Y chromosome ethnic introgression in a population of a given ethnic origin.

Superoxide dismutase, catalase and glutathione peroxidase activities in human blood: influence of sex, age and cigarette smoking.

OBJECTIVE: To obtain reference ranges for each of the main antioxidant enzymes (AOE) and analyze the influence of sex, age, and cigarette smoking on AOE activity in human blood. DESIGN AND METHODS: We investigated superoxide dismutase (SOD), catalase (CAT), and seleno-dependent glutathione peroxidase (GSH-Px) activities in the whole blood from 103 healthy subjects, from 18-67 years old (51 males and 52 females). RESULTS: We found a large and highly significant interindividual variability in the activity of all the AOE studied (p < 0.001). The interindividual coefficients of variation were 13.5% for SOD, 21.0% for CAT, and 36.2% for GSH-Px, indicating that GSH-Px exhibits the highest interindividual variability. Females showed higher SOD (p < 0.001) and CAT (p < 0.001) activities but lower GSH-Px (p < 0.05) activity than males. We found a significant effect of age on SOD activity (p < 0.001), showing that in human blood it decreases with age and that this decrease is not linear, beginning at 28 years of age. We also observed a linear effect of age on GSH-Px activity indicating that the activity of this enzyme increases with age (p < 0.01). No effect of age on CAT activity was observed (p > 0.05). AOE activity in smokers was found not to be significantly different from that observed in non-smokers (p > 0.05) except in the case of CAT activity in females, which was found to be lower in smokers than in non-smokers (p < 0.05). In addition, we determined reference ranges for the activity of each antioxidant enzyme studied. CONCLUSIONS: Our results confirm that AOE activity in human blood exhibits a wide interindividual variability and suggest that this variability may be ascribed, at least in part, to the sex and age of the individuals. Moreover, our results suggest that cigarette smoking does not influence AOE activity in human blood. Accordingly, it is suggested that for clinical purposes it may be necessary to consider the sex and age of the subjects involved in the study.

Effects of antioxidants on streptonigrin-induced DNA damage and clastogenesis in CHO cells.

The effect of several free radicals scavengers on DNA damage and clastogenesis induced by streptonigrin (SN) in CHO cells was investigated. The addition of the antioxidant enzymes superoxide dismutase and/or catalase on CHO cell cultures did not prevent the induction of DNA and chromosome damage by SN. In fact, when superoxide dismutase was added to the culture medium an increase on the frequency of SN-induced chromosome aberrations was observed. Moreover, the addition of the hydroxyl radicals scavenger mannitol caused a significant increase in DNA and chromosome damage induced by SN. On the contrary, when all the antioxidants mentioned above were added-alone or in different combinations-encapsulated into liposomes, a significant decrease in the yield of SN-induced chromosome aberrations and DNA damage was observed. These findings indicate that free radicals are involved in the production of DNA and chromosome damage by SN and that this damage can be partially inhibited through the incorporation of antioxidants by the cells.

Origin of Amerindian Y-chromosomes as inferred by the analysis of six polymorphic markers.

We analysed the frequency of six Y-specific polymorphisms in 105 Amerindian males from seven different populations, 42 Caucasian males, and a small number of males of African, Chinese, and Melanesian origin. The combination of three of the six polymorphisms studied produced four different Y-haplogroups. The haplogroups A (non-variant) was the most frequent one. Eighty-five percent of Amerindians showing haplogroup A have the alphoid II (alpha hII) and the DYS19A Y-specific markers, an association that is found only in 10% of Caucasians and that has not been detected in Asiatics and Africans. Haplogroups C (YAP+) and D (YAP+ plus an A-->G transmission in the locus DYS271) are of African origin. Four percent of Amerindians and approximately 12% of Caucasians showed haplogroup C; approximately 1% of Amerindians and approximately 2% of Caucasians had haplogroup D. Haplogroup B is characterized by a C-->T transition in nucleotide position 373 of the SRY gene domain; this haplogroup is found in Caucasians (approximately 12%) and Amerindians (approximately 4%). None of the Amerindians exhibiting the haplogroups B, C, or D show the haplotype alpha hII/DYS19A. By haplotyping the the Alu insert and the DNA region surrounding the insert in YAP+ individuals, we could demonstrate that Amerindian Y chromosomes bearing African markers (haplogroups C and D) are due to recent genetic admixture. Most non-alpha hII/DYS19A Amerindian Y-chromosomes in haplogroup A and most cases in haplogroup B are also due to gene flow. We show that haplotype alpha hII/DYS19A is in linkage disequilibrium with a C-->T transition in the locus DYS19A. Our results suggest that most Amerindian Y-chromosomes derive from a single paternal lineage characterized by the alpha hII/DYS19A/DYS199T Amerindian-specific haplotype. The analysis of a larger sample of native American Y-chromosome will be required in order to confirm or correct this hypothesis.

Worldwide distribution of human Y-chromosome haplotypes.

We surveyed several human populations worldwide with three PCR-based polymorphisms located in the human Y chromosome: the alphoid heteroduplex [alpha(h)] polymorphic system, the DYS19 microsatellite locus, and a polymorphic Alu insertion (YAP). By typing with the former two polymorphisms [alpha(h) and DYS19] we found 46 different haplotypes in 364 males from several populations worldwide. There were significant geographic differences in the distribution of the haplotypes, several of which were seen in only one population and can be used as populational markers in future surveys. The haplotypic diversity in major ethnic groups revealed high levels in Greater Asians, followed by Africans and Caucasians, and a very low diversity was seen in Amerindians. The discrimination probability of such haplotypes for a random sample of Brazilian Caucasians was 0.82, suggesting great potential usefulness in forensic studies. The parsimonious relationship between different alpha(h) types and the addition of the YAP polymorphism data allowed the construction of an informative picture of the origin and evolution of the antiquity of the alpha(h) polymorphism. The DYS19 allele diversity related to each alpha(h) type allowed a crude estimation of the antiquity of many alpha(h) types. These ancient alpha(h) types were present in different populations suggesting a common ancestor that could antedate the first out-of-Africa migrations.

Enhancement of chromosome aberrations by the combination of DNA substitution with halogenated deoxyuridine and streptonigrin treatments.

We treated CHO cells with streptonigrin (SN) alone, in combination with BrdUrd or IdUrd substitution, and with or without the addition of caffeine. The cells assessed for chromosome damage by SN were in the G2 period and the magnitude of the damage was expressed as monosubstituted chromatid breaks, bisubstituted chromatid breaks and boundary regions breaks (boundary regions indicate the point of exchange of mono- and bisubstituted chromatids). We found that the combination of BrdUrd or IdUrd substitution with SN treatments produced a remarkable increase in the frequency of breaks over the frequencies observed with the halogenated compound only. The effect was more evident with IdUrd than with BrdUrd, and more dramatic in bisubstituted than in monosubstituted chromatids. The frequency of boundary breaks in cells treated with BrdUrd plus SN was similar to the frequency of breaks in monosubstituted chromatids treated similarly. Conversely, the damage in boundary regions was almost similar to that in bisubstituted chromatids in cells challenged with IdUrd plus SN. The addition of caffeine to BrdUrd-substituted chromosomes gave rise to a marked enhancement of breakages with a gradient of chromatid damage that was: bisubstituted > monosubstituted > boundary regions. A further increase of chromatin breaks maintaining the gradient indicated above was obtained when the cells were treated with BrdUrd plus SN plus caffeine. We propose that BrdUrd and IdUrd substitution alone or in combination with caffeine treatments and with SN in its capacity to bind DNA, give rise to different chromatin structures capable of modulating the DNA damage induced along the chromatin fibril by the active oxygen species liberated by SN-DNA complexes.