Inter-individual variation in DNA double-strand break repair in human fibroblasts before and after exposure to low doses of ionizing radiation

Abstract

DNA double-strand breaks (DSB) are generally considered the most critical lesion induced by ionizing radiation (IR) and may initiate carcinogenesis and other disease. Using an immunofluorescence assay to simultaneously detect nuclear foci of the phosphorylated forms of histone H2AX and ATM kinase at sites of DSBs, we examined the response of 25 apparently normal and 10 DNA repair-deficient (ATM, ATR, NBN, LIG1, LIG4, and FANCG) primary fibroblast strains irradiated with low doses of ¹³⁷Cs γ-rays. Quiescent G₀/G₁-phase cultures were exposed to 5, 10, and 25 cGy and allowed to repair for 24 h. The maximum level of IR-induced foci (0.15 foci per cGy, at 10 or 30 min) in the normal strains showed much less inter-individual variation (CV ≈ 0.2) than the level of spontaneous foci, which ranged from 0.2–2.6 foci/cell (CV ≈ 0.6; mean ± SD of 1.00 ± 0.57). Significantly slower focus formation post-irradiation was observed in seven normal strains, similar to most mutant strains examined. There was variation in repair efficiency measured by the fraction of IR-induced foci remaining 24 h post-irradiation, curiously with the strains having slower focus formation showing more efficient repair after 25 cGy. Interestingly, the ranges of spontaneous and residual induced foci levels at 24 h in the normal strains were as least as large as those observed for the repair-defective mutant strains. The inter-individual variation in DSB foci parameters observed in cells exposed to low doses of ionizing radiation in this small survey of apparently normal people suggests that hypomorphic genetic variants in genomic maintenance and/or DNA damage signaling and repair genes may contribute to differential susceptibility to cancer induced by environmental mutagens.

Introduction

Ionizing radiation (IR) is a ubiquitous environmental mutagen and a prototypical DNA-damaging agent. Health effects of low doses of IR continue to be a matter of debate for concerns of radiological protection. Human population heterogeneity in radiosensitivity is dramatically illustrated by rare genetic disorders such as ataxia-telangiectasia (A-T), Nijmegen breakage syndrome (NBS), and ligase IV deficiency (LIG4 syndrome). Cells derived from these patients are hypersensitive to IR due to mutations impacting DNA double-strand break (DSB) recognition, signaling, and repair capacity [1], [2], [3]. Many studies employing cell survival and chromosome aberration assays have shown significant variation among primary fibroblast strains derived from apparently normal individuals in the general population [4], [5], [6], [7]. Interestingly, while strains derived from patients with A-T and other syndromes were more sensitive on average than normal strains, in each case the overall response of the mutant strains overlapped the distribution of the normal strains [5]. The majority of these in vitro studies employed relatively high doses of IR (≥50 cGy) because of limited experimental sensitivity at low doses. However, there is accumulating evidence of inter-individual variation for induced chromosomal aberrations after low IR doses [8], [9], [10] and for cell survival and γ-H2AX foci induction during continuous low dose-rate exposure [7], [11], [12].

Epidemiological studies indicate that cancer risk increases with IR exposure even at low doses [13], [14], [15], [16]. Reviews of individual and pooled studies of populations that received accidental, medical, or occupational IR exposures support a linear increase in cancer risk with increasing dose, with large uncertainties of risk at low dose. A study examining 55 single nucleotide polymorphisms (SNPs) in 17 DNA repair genes in a large cohort of U.S. radiologic technologists revealed a significant correlation between breast cancer risk, low dose occupational IR exposure, and four SNPs in BRCA1 and PRKDC [16]. Much work remains to determine how genetic variation in DNA repair and related genes [17], [18], [19], [20] impacts individual response and relative health risk after low-dose exposure.

Directly induced DSBs, as well as bi-stranded, clustered DNA damage that may be processed into DSBs [21], are considered the critical IR-induced lesions associated with cancer risk [22], [23], [24]. Studies of the activation of ATM and Tp53-dependent signaling pathways [25], [26] suggest that normal cells may exhibit a threshold for the recognition of DSBs induced by low doses. It is likely that incorrectly repaired DSBs underlie the chromosomal instability associated with neoplastic transformation as evidence of activated DNA damage response pathways and increased DSB levels have been observed in precancerous tissues [27], [28], [29].

Immunochemical detection of histone H2AX phosphorylated at Ser139 (γ-H2AX) provides a sensitive measure of DSB induction and repair in various cell types, including quiescent G₀/G₁-phase human fibroblasts after low dose IR [11], [12], [30], [31], [32], [33], [34], [35], [36], [37]. Activation of the protein kinase ATM by DSBs is accompanied by its intermolecular auto-phosphorylation at Ser1981 [38]. In irradiated fibroblasts ATM^S1981-P (pATM) foci form as early as 5 min after irradiation with 20 cGy [38], co-localize with γ-H2AX foci, and display similar kinetics of formation and disappearance [39]. We have used an assay for enumerating DSBs based on this co-localization [38], [39], [40], [41] to assess the degree of variation among quiescent G₀/G₁-phase primary fibroblasts derived from a small group of 25 apparently normal individuals for: (1) the level of spontaneous DSBs, (2) the yield and kinetics of recognition of IR-induced DSBs, and (3) the efficiency and dose dependence of DSB repair by NHEJ following doses of 5, 10, and 25 cGy of ¹³⁷Cs γ-rays. To gain a sense of the biological significance of the variation observed in the normal strains, we also assayed mutant fibroblast strains defective in DSB recognition, signaling, and repair: ATM [1], ATR [3], [42], NBN (NBS1) [2], [43], LIG1 [44], LIG4 [3], [45], [46], and FANCG [47].