Predisposition to cancer and radiosensitivity

Many cancer-prone diseases have been shown to be radiosensitive. The radiosensitivity has been attributed to pitfalls in the mechanisms of repair of induced DNA lesions or to an impaired cell cycle checkpoint response. Although discrepancies exist in the results obtained by various authors on the radiosensitivity of individuals affected by the same disease, these can be attributed to the large variability observed already in the response to radiation of normal individuals. To date three test are commonly used to assess radiosensitivity in human cells: survival, micronucleous and G2 chromosomal assay. The three tests may be performed using either fibroblasts or peripheral blood lymphocytes and all the three tests share large interindividual variability. In this regard a new approach to the G2 chromosomal assay which takes into account the eventual differences in cell cycle progression among individuals has been developed. This new approach is based on the analysis of G2 homogeneous cell populations. Cells irradiated are immediately challenged with medium containing bromodeoxyuridine (BrdUrd). Then cells are sampled at different post-irradiation times and BrdUrd incorporation detected on metaphases spread and the scoring is done only at time points showing similar incidence of labelled cells among the different donors. Using this approach it has been possible to reduce the interindividual variability of the G2 chromosomal assay.


INTRODUCTION: RADIOSENSITIVITY AND HUMAN SYNDROMES
Interest in the radiosensitivity of human cells has developed from the identification of a series of genetic disorders that not only exhibit large, unambiguous alterations in sensitivity to radiation and genotoxic chemicals but are also cancer-prone.
Xeroderma pigmentosum (XP) was the first disease to be associated with a major alteration in radiation response (Cleaver, 1968(Cleaver, , 1969)).XP patients exhibit a hereditary susceptibility to sunlight-induced cancer correlated with a genetic defect in the ability to repair DNA damage (Cleaver and Kraemer, 1989).Fibroblasts and lymphoid cells from XP patients showed hypersensitivity to ultraviolet (UV) light damage in vitro.In this view, many investigations of the relationship between radiation sensitivity and cancer susceptibility in a wide range of genetic disorders were performed, and it was demonstrated that several other diseases are also characterized by a hypersusceptibility to DNA-damaging agents, like Fanconi's anemia syndrome, sensitive to DNA cross-linking agents (Fujiwara and Tatsumi, 1977) and Werner's syndrome, susceptible to the carcinogen 4-nitroquinoline-oxide (Gebhart et al., 1985).
A deficiency in DNA repair was clearly demonstrated to be an underlying cause of UV-sensitivity in XP and Cockayne syndrome, but more complex abnormalities in the processing of DNA damage induced by X-irradiation were identified in ataxia telangiectasia (AT).
Evidence from these disorders sets the stage for the idea that a wide range of hereditary clinical disorders could be related in varying degrees to an increased sensitivity to DNA-damaging agents and radiosensitivity.
The most striking example is AT, an autosomal recessive disorder that causes immunological dysfunction, proneness to cancer, and an unusual susceptibility to Xirradiation (Higurashi and Conen, 1973; for a recent review see Lavin and Khanna, 1999).Also the Nijmegen syndrome, an autosomal recessive disease previously identified as AT variant, is characterized by a high sensitivity to X-irradiation (see Digweed et al., 1999).It has also been demonstrated that many other syndromes or inherited conditions are characterized by increased radiosensitivity (Table I).Among them, both hereditary syndromes characterized by wide genomic instability (e.g., Bloom's syndrome) and cancer syndromes (Wilm's tumor).
It is noteworthy that almost all these pathological situations are reported to indicate predisposition towards cancer in the affected individuals.For this reason, it could be very interesting to assess whether a strong correlation between radiosensitivity and predisposition to cancer exists.
A typical example are AT heterozygotes which are more radiosensitive than healthy controls and have an ap-

Abstract
Many cancer-prone diseases have been shown to be radiosensitive.The radiosensitivity has been attributed to pitfalls in the mechanisms of repair of induced DNA lesions or to an impaired cell cycle checkpoint response.Although discrepancies exist in the results obtained by various authors on the radiosensitivity of individuals affected by the same disease, these can be attributed to the large variability observed already in the response to radiation of normal individuals.To date three test are commonly used to assess radiosensitivity in human cells: survival, micronucleous and G 2 chromosomal assay.The three tests may be performed using either fibroblasts or peripheral blood lymphocytes and all the three tests share large interindividual variability.In this regard a new approach to the G 2 chromosomal assay which takes into account the eventual differences in cell cycle progression among individuals has been developed.This new approach is based on the analysis of G 2 homogeneous cell populations.Cells irradiated are immediately challenged with medium containing bromodeoxyuridine (BrdUrd).Then cells are sampled at different post-irradiation times and BrdUrd incorporation detected on metaphases spread and the scoring is done only at time points showing similar incidence of labelled cells among the different donors.Using this approach it has been possible to reduce the interindividual variability of the G 2 chromosomal assay.
In contrast, about 40% of an unselected group of breast cancer cases were found to be radiosensitive (Scott et al., 1994).This finding suggests that radiosensitivity could actually be a potential predisposing condition to breast cancer (Roberts et al., 1999).In the study of Scott and colleagues (1994), the authors also found that about 9% of healthy controls were radiosensitive.
Many other studies were performed to determine the incidence of radiosensitive individuals in the general population.For example, Sanford and colleagues (1989) reported that about 5% of the normal population is radiosensitive.
However, discrepancies exist in the results presented in each study.In particular, conditions proved to be radiosensitive by one author are not by others (Scott et al., 1996;Palitti et al., 1999).One of the recognized problems is the variability in the response to radiation of apparently normal individuals.This problem was extensively in-vestigated by Little and colleagues (1989), who demonstrated that great variability exists in the response to radiation of normal cells.This finding complicates the use of radiosensitivity as a marker of cancer predisposition.In fact, this variability makes it very difficult to determine the range in which the individuals are to be considered radiosensitive or normal.Moreover, it is not proven that these are always cancer-prone conditions and thus all the individuals predisposed to cancer have to be radiosensitive.In fact, radiosensitivity would result, from defects in the repair of radiation-induced DNA damage or defects in the processing and/or signalling of this damage.
In this regard, a lot of genes could be involved in the radiosensitive phenotype, only a few of which are known (Table II) and it is unlikely that all these genes can be correlated with predisposition to cancer.
On the other hand, apparently normal radiosensitive individuals could be examined to discover new genes implicated in the radiation response.

PATHWAYS INVOLVED IN THE REPAIR OF IONIZING RADIATION-INDUCED DOUBLE-STRAND BREAKS
Double-strand breaks (DSB) are generated by endogenously produced radicals and exogenous agents such as ionizing radiation (IR), which is often used in anti-cancer therapy.Repair of DSBs is of cardinal importance to prevent chromosomal fragmentation, translocations and deletions.In the soma, the genomic instability resulting from persistent or incorrectly repaired DSBs can lead to carcinogenesis through activation of oncogenes, inactivation of tumor-suppressor genes or loss of heterozygosity, while in the germline they can lead to inborn defects (see Kanaar et al., 1998).
The deleterious effects of DSBs have triggered the evolution of multiple pathways for their repair (see Kanaar et al., 1998;Critchlow and Jackson, 1998;Pastink and Lohman, 1999).
i) single-strand annealing (SSA), ii) non-homologous end-joining (NHEJ), iii) recombinational repair (RR) see Figure 1.All three mechanisms are conserved in evolution but their relative contributions differ between higher and lower eukaryotes.The relative importance of each pathway also depends on the phase of the cell cycle, NHEJ being mostly active during G1/early-S but RR and SSA active only during late-S/G2 (Takata et al., 1998).
Homologous recombination requires extensive regions of DNA homology and repairs DSBs accurately using information on the undamaged sister chromatid or homologous chromosome.This pathway depends upon the presence of the protein products of the RAD52 epistasis group of genes.These proteins, which appear conserved from yeast to man, essentially act as during meiotic crossing-over (see Figure 1; Haber, 2000).
DNA end-joining, on the other hand, uses no, or ex- tremely limited, sequence homology to rejoin juxtaposed ends in a manner that need not be error free (Figure 1).This repair system, first identified in mammals, has also been found in yeast; it repairs predominantly through RR, and it is dependent upon the activity of the DNA-PK complex and that of XRCC4/Ligase 4 (Jeggo et al., 1999).
The SSA pathway was proposed initially to explain the results of intramolecular recombination in plasmid DNAs introduced into mammalian cells or X. laevis oocytes.Later on it became evident that repair by SSA also occurs in yeast.The SSA pathway is dependent on the presence of direct repeats on both sides of the break and thus does not require extensive homology.This pathway depends upon the presence of the RAD52 protein as well as of the mismatch repair enzyme MSH2 and MSH3, at least in yeast (Pastink and Lohman, 1999).
Not only are the enzymes of the DSB repair systems important for determining the cellular sensitivity to ionizing radiation but also those involved in the signalling of the DNA damage to the checkpoint mechanism.This is proved by the radiosensitive phenotype of the human syndrome ataxia telangiectasia, in which large defects in the checkpoint activation are present.(See Table II for a summary of the properties of the proteins involved in the response to IR-induced DSBs and the corresponding cellular phenotypes in mammalian cells).
The common use of IR in anti-tumor radiotherapy, as well as the observation of links between radiosensitivity and cancer proneness, make it paramount to give attention into the molecular intricacies of IR-induced DNA damage repair when looking for a radiosensitive phenotype.In the past few years, the availability of known genes has provided the essential tools to translate the results from the initial genetic analyses to molecular mechanisms of DSB repair.
Currently, a major challenge ahead is the evaluation of the relative contribution of the different DSB repair pathways survival after exposure to IR in different mammalian tissues and tumors in the view of assessing the risks arising from radiation exposure.

RADIOSENSITIVITY ASSAYS
To date, three tests are commonly used to assess radiosensitivity in human cells: ♦ Survival assay; ♦ Micronucleus assay; ♦ G2 chromosomal assay.
The first uses the survival after irradiation with different doses to obtain and compare the D0 values (D0 values are the dose value at which the survival is reduced to 37% of the untreated culture).A low D0 value should correspond to a radiosensitive phenotype.
The second looks at the X-ray-induced micronuclei, both after low and high irradiation rates (giving better reproducible results according to Scott et al., 1998).Higher yields of micronuclei correlate with radiosensitivity.
The third uses the induction of chromatid-type aberrations after irradiation (50 cGy usually) in the G2-phase of the cell cycle.Also in this case, higher yields of chromosomal damage correlate with radiosensitivity.
The three tests may be performed using either fibroblasts or peripheral blood lymphocytes, the former being preferred for the survival assay.All three tests share the interindividual variability described above.Reasons for this variability have been attributed to differences in the experimental procedures used by different investigators (Scott et al., 1996) or to differences in cellular growth among the individuals (Darroudi et al., 1995;Palitti et al., 1999).For example, in the survival assay, differences in the relative duration of G1, S and G2 phases of the cell cycle, which are not uniform in their sensitivities could influence radiation sensitivity.On the other hand, different responses in the duration of the G2 arrest after irradiation could give a large variability in the yield of chromosomal damage using the G2 assay.
In this regard, a new approach to the G2 assay, which takes into account these differences in the extent of the X-ray-induced G2 delay has been developed by Palitti and colleagues (1999).This new approach is based on the analysis of the G2 homogeneous population in mitosis.Cells are irradiated and immediately challenged in medium containing bromodeoxyuridine (BrdUrd).Then, cells are harvested at different post-irradiation times and BrdUrd incorporation is visualized by means of immunocytochemical detection on metaphase spreads.For each time-point the percentage of labelled cells in mitosis is calculated and chromosomal damage scored in unlabelled mitosis (G2) only at time-points showing a similar incidence of labelled cells (usually around 10%).Using this approach, the authors have actually been able to reduce the interindividual variability of the G2 assay.

CONCLUDING REMARKS
It may be useful to develop a simple cytogenetic assay which would allow the screening of a large number of individuals for radiosensitivity in order to direct the positive cases to more relevant screening for cancer susceptibility.In addition, unambiguous identification of radiosensitive patients with cancer could be of importance for the correct use of the radiotherapy in oncology.Both situations require a good test for radiosensitivity.

Table II -
Genes involved in the repair and processing of IR-induced DNA damage.