Molecular and Cellular Responses to Ionization Radiation in Untransformed Fibroblasts from the Rothmund–Thomson Syndrome: Inﬂuence of the Nucleo-Shuttling of the ATM Protein Kinase

. Simple Summary


Introduction
Rothmund-Thomson syndrome is a rare disease discovered jointly by a German ophthalmologist, Auguste Rothmund, and an English dermatologist, Sydney Thomson. In 1868, Rothmund associated skin rash and juvenile cataracts in some of his patients [1]. From 1921, Thomson named "poikiloderma congenitale" all of the erythema associated with the symptoms that Auguste Rothmund had described but also noticed that the erythema can  [13] Baller-Gerold RECQL4 RECQL4 helicase RTS symptoms with Craniosynostosis; Limb abnormality Thumb/radial hypoplasia [14] While the RTS patients may be exposed to ionizing radiation (IR) through radiodiagnosis for their clinical follow-up and eventually through radiotherapy to treat their tumor (for the RTS2 patients), the question of their radiosensitivity was raised in the 1980s: some authors investigated the radiobiological features of lymphocytes and fibroblasts from the RTS patients (Table 2 and references herein). In vitro clonogenic survival and cytogenetic features taken as endpoints or observations of post-radiotherapy reactions in healthy tissues suggest that RTS may be associated with a slight to moderate radiosensitivity. In parallel, a number of studies has aimed to elucidate the biological role of the RECQL4 protein in the DNA damage and signaling pathways, notably the management of the DNA double-strand breaks (DSB), the key-damage of the radiation-induced lethal effect [11,[20][21][22][23][24]. However, most of these studies involved tumor or transformed fibroblast cell lines with very few cells from RTS donors. However, to reliably quantify the radiation-induced (RI) risks, i.e., the individual radiosensitivity (risk of post-radiotherapy tissue over-reactions), the radiosusceptibility (risk of RI cancer), and the radiodegeneration (risk of RI aging) of the RTS patients, represents a societal, medical, and scientific issue [25]. For example, in France, to evaluate the risks linked to the medical exposure to IR is inasmuch important as a recent governmental decree that recommends radiologists take into account the individual radiosensitivity and radiosusceptibility of their patients to justify irradiating acts [26]. Table 2. Non-exhaustive list of studies dealing with the radiobiological characterization of RTS cells.

fibroblasts cell lines
Cell survival DNA repair Repair replication Significant radiosensitivity [27] 1 fibroblast cell line Cell survival Slight radiosensitivity [28] 1 fibroblast cell line Colony formation Enhanced Radiosensitivity But not toxicity to drugs [29] 6 fibroblast cell lines Colony formation 1.2 fold radiosensitivity [30] 2 lymphocytes G2 assay Chromosomal radiosensitivity [31] 1 patient Clinical radiosensitivity after radiotherapy [32] 1 patient Clinical radiosensitivity after radiotherapy [33] 3 fibroblast cell lines from RTS patients 2 fibroblast cell lines from RTS patients' parents Despite considerable efforts and a large spectrum of predictive assays, the quantification of the RI risks is still not consensual [34]. Since 2014, we have proposed a general and unified mechanistic model of the individual response to IR based on the RI ATM nucleo-shuttling (RIANS). The applications of the RIANS are numerous and concern both high-and low-dose phenomena. Particularly, the RIANS model may serve to predict radiosensitivity in a number of exposure conditions and was found relevant for more than twenty genetic diseases (e.g., [35][36][37][38]). To date, the RIANS model provides the highest statistically robust prediction of post-radiotherapy radiosensitivity [39][40][41][42][43] and a unified model to explain the linear-quadratic formula that links cell survival and dose [44]. More recently, we provided proof that assays using the RIANS biomarkers as endpoints are the most powerful molecular approaches to evaluate radiosensitivity with the clonogenic cell survival assay [45].
Here, by using successively clonogenic cell survival, micronuclei, and immunofluorescence assays with the RIANS biomarkers, we applied a routine procedure of radiobiological characterization [35,36,38,[46][47][48][49][50][51] to document the individual molecular and cellular responses of fibroblast cell lines derived from three RTS patients and two of their relatives.

Cell Lines
All of the experiments were performed with untransformed fibroblast cells in the plateau phase of growth to avoid any cell cycle-dependent artifacts. Routine cell cytometry data indicated that 90-99% of cells were in G0/G1 phase [39,52]. Hence, the activity of the homologous recombination pathway during the experiments described here can be considered negligible [25]. Three radioresistant (1BR3, MRC5, and Hs27) controls that originated from radioresistant and apparently healthy patients and the hyperradiosensitive ATM-mutated (AT4BI) and LIG4-mutated (180BR) cell lines were used in this study as negative and positive controls for radiosensitivity. These cell lines belong to the COPERNIC collection that has been abundantly documented and is composed of radioresistant, hyper-radiosensitive gifted cell lines and fibroblast cell lines derived from RT-treated patients. The COPERNIC database is protected under the reference IDDN.FR.001.510017.000.D.P.2014.000.10300. All sampling protocols of the COPERNIC collection were approved by the national ethical committee in agreement with the current national regulations. The resulting cells were declared under the numbers DC2008-585, DC2011-1437, and DC2021-3957 to the Ministry of Research. Three fibroblast cell lines from RTS patients were used (81CLB from the COPERNIC collection and AG17524 and AG18371 purchased from the Coriell Cell Repositories (Camden, NJ, USA)). Lastly, two fibroblast cell lines from RTS donor patients (AG18466 and AG18373), also purchased from the Coriell Cell Repositories, were also used in this study. Table 3 provides the major clinical and genetic features of all cell lines tested here.

Clonogenic Cell Survival Assay
The intrinsic cellular radiosensitivity was quantified from clonogenic cell survival data obtained from standard delayed plating procedures [54]. The survival data were fitted to the linear-quadratic (LQ) model that describes the cell survival S as a function of dose D, as follows: where α and β are adjustable parameters to be determined [44].

Micronuclei Assay
During each immunofluorescence experiment, the DAPI counterstaining permits quantification of the micronuclei [55]. With such an approach, the micronuclei yield assessed is not numerically equivalent to that obtained with the micronucleus assay involving cytochalasin B [56]. Micronuclei data were expressed as the number of micronuclei per 100 cells (nuclei).

Statistical Analysis
Two-way ANOVA was used to compare two numerical values, and Spearman's test was used to compare the kinetic data. The foci kinetic data were fitted to the so-called Bodgi's formula that describes the kinetics for the appearance/disappearance of nuclear foci formed by some protein relocalization after irradiation [58]. Statistical analysis was performed by using Kaleidagraph v4 (Synergy Software, Reading, PA, USA) and Graphpad Prism (San Diego, CA, USA).

Abnormally High Levels of Micronuclei in RTS Fibroblasts
The propagation of unrepaired DNA breaks up to mitosis may lead to irreversibly damaged chromosomal fragments causing mitotic death. Some chromosomal fragments may escape from the metaphases and generate micronuclei. Micronuclei have been shown to be quantitatively correlated with cellular radiosensitivity [25]. No significant difference was found between the spontaneous micronuclei in RTS cells and those in radioresistant controls (p > 0.5). Conversely, the number of spontaneous micronuclei per 100 cells was significantly higher in the two hyper-radiosensitive fibroblasts (p < 0.02). After 24 h postirradiation, the number of residual micronuclei per 100 cells was significantly higher in the RTS fibroblasts (p < 0.001) compared with the radioresistant controls but lower than that observed in the hyper-radiosensitive fibroblasts (p < 0.0001) (Figure 2A), supporting a significant genomic instability in the RTS fibroblasts tested (Figure 2A). The average number of residual micronuclei per 100 cells was 13.3 ± 2.0 in the RTS fibroblasts while it was 55 ± 10 and 1.1 ± 0.5 in the hyper-radiosensitive and radioresistant control fibroblasts, respectively. The survival and micronuclei data from the three fibroblast cell lines were in full agreement with the correlation established between SF2 and micronuclei with fibroblasts from different genetic diseases and published recently [45] ( Figure 2B).

Abnormal Number of γH2AX Foci in RTS Fibroblasts after Irradiation
The micronuclei data suggest that a significant yield of DSB may remain unrepaired and participates in the lethal effect. To investigate the recognition and repair of radiationinduced DSB via the non-homologous end-joining (NHEJ) pathway, the most predominant DSB repair pathway in humans, we applied immunofluorescence to the phosphorylated form of the H2AX variant histone (γH2AX). According to the RIANS model, in response to radiation, cytoplasmic ATM dimers monomerize. The ATM monomers diffuse in the nucleus and phosphorylate the H2AX molecules at the DSB sites early after irradiation [43]. The ATM-dependent phosphorylation of H2AX histones is responsible for the formation of γH2AX foci reflecting the DSB recognized by NHEJ. Hence, the early γH2AX foci indicate the DSB recognition by NHEJ while the residual γH2AX foci reflect the unrepaired DSB among the DSB recognized by NHEJ [43]. Some spontaneous γH2AX foci were observed in the three fibroblast cell lines, but their number was not significantly different from that in the radioresistant controls (p > 0.5). In the radioresistant controls, the number of γH2AX foci scored 10 min after 2 Gy was 79 ± 4 per cell, consistent with the value 37 ± 4 per Gy per cell obtained in many situations [47]. The number of induced γH2AX foci scored 10 min post-irradiation in the three fibroblast cell lines was significantly lower than that in the radioresistant controls (81CLB: 40 ± 8; AG17524: 25 ± 4; AG18371: 22 ± 4 γH2AX foci per cell; p < 0.001) ( Figure 3A). These findings suggest an impairment in the DSB recognition by NHEJ. It is noteworthy that the ATM-mutated AT4BI fibroblasts did not show any γH2AX foci while the LIG4-mutated 180BR fibroblasts elicited a slow γH2AX foci repair kinetic, reflecting a gross DSB repair defect, in full agreement with literature [37,39,45,47]. It must be stressed here that while ATR and DNA-PK kinases may phosphorylate H2AX, ATRand DNA-PKcs-mutated cells show γH2AX foci early after irradiation whereas the great majority of ATM-mutated cells generally do not [60]. For example, the AT5BI cell lines may show less than 10 γH2AX foci at 10 min post-irradiation. Both AT4BI and AT5BI have been shown elsewhere [37,48].
At 24 h post-irradiation, the RTS fibroblasts showed a similar number of γH2AX foci as the radioresistant controls (p > 0.3), suggesting that the repair rate of DSB recognized by NHEJ can be considered normal at this time. After 24 h post-irradiation, the number of DSB recognized by NHEJ and remaining unrepaired was similar in the radioresistant control and in the RTS patients' cells ( Figure 3A). Together, these data suggest a lack of DSB recognition by NHEJ that was not accompanied by a gross DSB repair defect of DSB recognized by NHEJ as in the LIG4-mutated cells. The fibroblasts from the parents of the RTS patients showed γH2AX foci kinetics similar to those observed in radioresistant controls (p > 0.8) ( Figure 3B).

Abnormal Number of γH2AX Foci in RTS Fibroblasts after Irradiation
The micronuclei data suggest that a significant yield of DSB may remain un and participates in the lethal effect. To investigate the recognition and repair of r induced DSB via the non-homologous end-joining (NHEJ) pathway, the most p nant DSB repair pathway in humans, we applied immunofluorescence to the ph lated form of the H2AX variant histone (γH2AX). According to the RIANS mod sponse to radiation, cytoplasmic ATM dimers monomerize. The ATM monomer in the nucleus and phosphorylate the H2AX molecules at the DSB sites early aft ation [43]. The ATM-dependent phosphorylation of H2AX histones is responsib formation of γH2AX foci reflecting the DSB recognized by NHEJ. Hence, the early foci indicate the DSB recognition by NHEJ while the residual γH2AX foci reflec

Abnormal Number of pATM Foci after Irradiation in RTS Fibroblasts
According to the RIANS model, since the formation of γH2AX foci is ATM-dependent, a low number of early γH2AX foci suggests that the diffusion of ATM in the nucleus is delayed. Once in the nucleus, the ATM monomers reassociate after the DSB repair process to form ATM dimers, easily quantifiable as nuclear foci by immunofluorescence with the specific antibodies of the auto-phosphorylated forms of ATM (pATM) [43]. The number of pATM foci assessed after 2 Gy was 40 ± 4 in the radioresistant controls, in agreement with previous studies ( Figure 4A). In the RTS fibroblasts, the number of pATM foci per cell observed 10 min post-irradiation was significantly lower than that in control fibroblasts (81CLB: 22 ± 6; AG17524: 19 ± 4; AG18371: 19 ± 3; p < 0.001). A similar conclusion was reached with pATM foci data assessed at 1 h post-irradiation (p < 0.05). In agreement with the literature, the ATM-mutated fibroblasts did not show pATM foci while the LIG4mutated showed normal pATM foci kinetics. These data suggest that the RTS fibroblasts elicited abnormal ATM nuclear kinase activity in response to radiation ( Figure 4A). Similar to γH2AX foci, the fibroblasts derived from the parents of the RTS patients elicited normal pATM foci kinetics (p > 0.7) ( Figure 4B). ment with literature [37,39,45,47]. It must be stressed here that while ATR and DNA-PK kinases may phosphorylate H2AX, ATR-and DNA-PKcs-mutated cells show γH2AX foci early after irradiation whereas the great majority of ATM-mutated cells generally do not [60]. For example, the AT5BI cell lines may show less than 10 γH2AX foci at 10 min postirradiation. Both AT4BI and AT5BI have been shown elsewhere [37,48]. foci was plotted against the post-irradiation time. Each plot represents the mean ± SEM of three replicates. For clarity, only the 1BR3 data, the 81CLB, AG17524, and AG18371 data, the 180BR data, and the AG18466 and AG18373 data were fitted. Data were fitted to the Bodgi's formula [58]. The insert in panel A shows a representative example of γH2AX foci assessed 24 h post-irradiation in one RTS fibroblast cell line (81CLB) and one radioresistant control (1BR3), with the corresponding DAPI counterstaining. The white bar corresponds to 5 µm.
At 24 h post-irradiation, the RTS fibroblasts showed a similar number of γH2AX foci as the radioresistant controls (p > 0.3), suggesting that the repair rate of DSB recognized by NHEJ can be considered normal at this time. After 24 h post-irradiation, the number of DSB recognized by NHEJ and remaining unrepaired was similar in the radioresistant control and in the RTS patients' cells ( Figure 3A). Together, these data suggest a lack of DSB recognition by NHEJ that was not accompanied by a gross DSB repair defect of DSB recognized by NHEJ as in the LIG4-mutated cells. The fibroblasts from the parents of the RTS patients showed γH2AX foci kinetics similar to those observed in radioresistant controls (p > 0.8) ( Figure 3B).

Abnormal Number of pATM Foci after Irradiation in RTS Fibroblasts
According to the RIANS model, since the formation of γH2AX foci is ATM-dependent, a low number of early γH2AX foci suggests that the diffusion of ATM in the nucleus is delayed. Once in the nucleus, the ATM monomers reassociate after the DSB repair process to form ATM dimers, easily quantifiable as nuclear foci by immunofluorescence with the specific antibodies of the auto-phosphorylated forms of ATM (pATM) [43]. The number of pATM foci assessed after 2 Gy was 40 ± 4 in the radioresistant controls, in agreement with previous studies ( Figure 4A). In the RTS fibroblasts, the number of pATM foci per foci was plotted against the post-irradiation time. Each plot represents the mean ± SEM of three replicates. For clarity, only the 1BR3 data, the 81CLB, AG17524, and AG18371 data, the 180BR data, and the AG18466 and AG18373 data were fitted. Data were fitted to the Bodgi's formula [58]. The insert in panel A shows a representative example of γH2AX foci assessed 24 h post-irradiation in one RTS fibroblast cell line (81CLB) and one radioresistant control (1BR3), with the corresponding DAPI counterstaining. The white bar corresponds to 5 µm.
The survival and pATM data from the three fibroblast cell lines were found in full agreement with the correlation established between SF2 and early pATM foci with fibroblasts from different genetic diseases and published recently [45] (Figure 5). Such findings support again the quantitative consistence of our data.

Abnormal Number of MRE11 Foci after Irradiation in RTS Fibroblasts
According to the RIANS model, in addition to the formation of the γH2AX foci, the radiation-induced diffusion of ATM monomers in the nucleus also triggers the phosphorylation of the MRE11 nuclease protein, responsible for its inhibition, that is accompanied by the formation of nuclear MRE11 foci. We investigated therefore the occurrence of MRE11 foci in the RTS fibroblasts. In the radioresistant controls, the MRE11 foci appeared from 2 to 8 h post-irradiation and reached their maximal yield at 4 h (7 ± 2 MRE11 foci per cell), in agreement with previous data ( Figure 6A). The shape of the MRE11 foci kinetics of the three fibroblast cell lines appeared to be clearly different from that of the radioresistant controls (p < 0.01). For the 81CLB and AG18371 fibroblasts, the maximal number of MRE11 foci was reached at 1 h and decreased up to 0 at 24 h post-irradiation. For the AG17524 fibroblasts, the number of MRE11 increased progressively to reach its maximal value at 24 h post-irradiation. It is noteworthy that, for both ATMand LIG4-mutated fibroblasts, the formation of the MRE11 foci was impaired ( Figure 6A). Like for γH2AX and pATM foci, the fibroblasts deriving from the parents of the RTS patients elicited normal MRE11 foci kinetics (p > 0.7) ( Figure 6B). was reached with pATM foci data assessed at 1 h post-irradiation (p < 0.05). In agreement with the literature, the ATM-mutated fibroblasts did not show pATM foci while the LIG4mutated showed normal pATM foci kinetics. These data suggest that the RTS fibroblasts elicited abnormal ATM nuclear kinase activity in response to radiation ( Figure 4A). Similar to γH2AX foci, the fibroblasts derived from the parents of the RTS patients elicited normal pATM foci kinetics (p > 0.7) ( Figure 4B). , the hyper-radiosensitive LIG4-mutated 180BR (red circles) and the three RTS (orange symbols: 81CLB, circles; AG17524, triangles; AG18371 inverted triangles) fibroblast cell lines and to (B): 2 fibroblast cell lines from parents of RTS patients (blue symbols: AG18466, circles and AG18373, triangles) in addition to the radioresistant 1BR3 controls (green squares). The number of pATM foci was plotted against the post-irradiation time. Each plot represents the mean ± SEM of three replicates. For clarity, only the 1BR3 data, the 81CLB, AG17524, and AG18371 data, the 180BR, and the AG18466 and AG18373 data were fitted. Data were fitted to the Bodgi's formula [58]. The insert in panel A shows a representative example of pATM foci assessed 1 h post-irradiation in one RTS fibroblast cell line (81CLB) and one radioresistant control (1BR3), with the corresponding DAPI counterstaining. The white bar corresponds to 3 µm.
The survival and pATM data from the three fibroblast cell lines were found in full agreement with the correlation established between SF2 and early pATM foci with fibroblasts from different genetic diseases and published recently [45] (Figure 5). Such findings support again the quantitative consistence of our data. For clarity, only the 1BR3 data, the 81CLB, AG17524, and AG18371 data, the 180BR, and the AG18466 and AG18373 data were fitted. Data were fitted to the Bodgi's formula [58]. The insert in panel (A) shows a representative example of pATM foci assessed 1 h post-irradiation in one RTS fibroblast cell line (81CLB) and one radioresistant control (1BR3), with the corresponding DAPI counterstaining. The white bar corresponds to 3 µm.

Subcellular Localization and Expression of the RECQL4 Protein in RTS Fibroblasts
A delayed RIANS is generally caused by the interaction between ATM monomers and some ATM substrates that are overexpressed in the cytoplasm [43]. We therefore examined the existence of potential cytoplasmic ATM-RECQL4 complexes in RTS fibroblasts. To this aim, as a first step, we investigated the expression of the RECQL4 protein in both total and cytoplasmic cell extracts by using immunoblots. By considering the total protein extracts, the level of RECQL4 expression appeared higher in the three fibroblast cell lines than in the radioresistant controls. Interestingly, the cytoplasmic RECQL4 forms were found absent in radioresistant controls, suggesting that the RECQL4 protein is essentially nuclear in these cells, in agreement with the literature (Table 1 and references therein) (Figure 7 and supplementary file). Hence, immunoblots highlighted the existence of abundant cytoplasmic forms of the RECQL4 protein in the three fibroblast cell lines tested. Interestingly, cytoplasmic forms of the RECQL4 proteins also appeared abundant in heterozygous RTS carriers, suggesting that heterozygous RECQL4 mutations may also lead to over-expressed cytoplasmic RECQL4 proteins (Figure 7).
As a second step, the proximity ligation assay (PLA) was thereafter applied to the three RTS and one radioresistant control fibroblasts (Figure 8). The PLA assay revealed 2-3 times significantly more cytoplasmic ATM-RECQL4 complexes in the RTS fibroblasts (p < 0.01) than in radioresistant controls. The irradiation did not significantly influence the number of PLA foci reflecting the ATM-RECQL4 complexes (Figure 8).

Abnormal Number of MRE11 Foci after Irradiation in RTS Fibroblasts
According to the RIANS model, in addition to the formation of the radiation-induced diffusion of ATM monomers in the nucleus also trigg ylation of the MRE11 nuclease protein, responsible for its inhibition, tha by the formation of nuclear MRE11 foci. We investigated therefore t MRE11 foci in the RTS fibroblasts. In the radioresistant controls, the MRE from 2 to 8 h post-irradiation and reached their maximal yield at 4 h (7 ± cell), in agreement with previous data ( Figure 6A). The shape of the MR of the three fibroblast cell lines appeared to be clearly different from th sistant controls (p < 0.01). For the 81CLB and AG18371 fibroblasts, the ma MRE11 foci was reached at 1 h and decreased up to 0 at 24 h post-irr AG17524 fibroblasts, the number of MRE11 increased progressively to r value at 24 h post-irradiation. It is noteworthy that, for both ATM-and broblasts, the formation of the MRE11 foci was impaired ( Figure 6A). Lik pATM foci, the fibroblasts deriving from the parents of the RTS patien MRE11 foci kinetics (p > 0.7) ( Figure 6B).  [45], the SF2 and the corresponding γH2AX foci data assessed 24 h post-irradiation and obtained from 36 COPERNIC cell lines were plotted together (closed black circles). Each point corresponds to the mean ± standard error of the mean (SEM) of 3 independent triplicates, at least. The survival and pATM data from the RTS fibroblast cell lines shown in Figures 1 and 4A were added to the graph (red circles). The best data fit was obtained with the linear law: SF2 (pATMmax) = 1.422 × pATMmax; r 2 = 0.87 (dotted line).

Subcellular Localization and Expression of the RECQL4 Protein in RTS Fibroblasts
A delayed RIANS is generally caused by the interaction between ATM monomers and some ATM substrates that are overexpressed in the cytoplasm [43]. We therefore examined the existence of potential cytoplasmic ATM-RECQL4 complexes in RTS fibroblasts. To this aim, as a first step, we investigated the expression of the RECQL4 protein in both total and cytoplasmic cell extracts by using immunoblots. By considering the total protein extracts, the level of RECQL4 expression appeared higher in the three fibroblast Figure 6. Kinetics of MRE11 foci in RTS fibroblasts. Anti-MRE11 immunofluorescence was applied to (A): the radioresistant controls (green symbols: 1BR3, squares; MRC5, diamonds; Hs27, inverted triangles), the hyper-radiosensitive ATM-mutated AT4BI (red squares) and LIG4-mutated 180BR (red circles), and the three RTS (orange symbols: 81CLB, circles; AG17524, triangles; AG18371 inverted triangles) fibroblast cell lines and to (B): 2 fibroblast cell lines from the parents of RTS patients (blue symbols: AG18466, circles and AG18373, triangles) in addition to the radioresistant 1BR3 controls (green squares). The number of MRE11 foci was plotted against the post-irradiation time. Each plot represents the mean ± SEM of three replicates. Data were fitted to Bodgi's formula [58]. The insert in panel A shows a representative example of MRE11 foci assessed 1 h post-irradiation in one RTS fibroblast cell line (81CLB) and one radioresistant control (1BR3), with the corresponding DAPI counterstaining. The white bar corresponds to 3 µm. sentially nuclear in these cells, in agreement with the literature (Table 1 and references therein) (Figure 7 and supplementary file). Hence, immunoblots highlighted the existence of abundant cytoplasmic forms of the RECQL4 protein in the three fibroblast cell lines tested. Interestingly, cytoplasmic forms of the RECQL4 proteins also appeared abundant in heterozygous RTS carriers, suggesting that heterozygous RECQL4 mutations may also lead to over-expressed cytoplasmic RECQL4 proteins (Figure 7).  As a second step, the proximity ligation assay (PLA) was thereafter applied to the three RTS and one radioresistant control fibroblasts (Figure 8). The PLA assay revealed 2-3 times significantly more cytoplasmic ATM-RECQL4 complexes in the RTS fibroblasts (p < 0.01) than in radioresistant controls. The irradiation did not significantly influence the number of PLA foci reflecting the ATM-RECQL4 complexes (Figure 8).

Statin and Bisphosphonate Treatment Protects RTS Fibroblasts from Radiation
The combination of pravastatin and zoledronate treatments (ZOPRA) was shown to accelerate the RIANS and, in some genetic diseases, to decrease the radiosensitivity [35][36][37][38]. The ZOPRA treatment was therefore applied to the RTS fibroblasts with micronuclei and early γH2AX and pATM foci. With regard to micronuclei, the ZOPRA treatment significantly reduced the number of residual micronuclei only in the RTS 81CLB fibroblasts.

Statin and Bisphosphonate Treatment Protects RTS Fibroblasts from Radiation
The combination of pravastatin and zoledronate treatments (ZOPRA) was shown to accelerate the RIANS and, in some genetic diseases, to decrease the radiosensitivity [35][36][37][38]. The ZOPRA treatment was therefore applied to the RTS fibroblasts with micronuclei and early γH2AX and pATM foci. With regard to micronuclei, the ZOPRA treatment significantly reduced the number of residual micronuclei only in the RTS 81CLB fibroblasts. With regard to the early γH2AX and pATM foci, the ZOPRA treatment significantly increased the number of foci in all three fibroblast cell lines tested (Figure 9). Together, these data suggest that the ZOPRA treatment may protect the RTS fibroblasts against IR by accelerating the RIANS and stimulating the ATM kinase activity in the nucleus, at least partially.
Radiation 2023, 3, FOR PEER REVIEW 14 With regard to the early γH2AX and pATM foci, the ZOPRA treatment significantly increased the number of foci in all three fibroblast cell lines tested ( Figure 9). Together, these data suggest that the ZOPRA treatment may protect the RTS fibroblasts against IR by accelerating the RIANS and stimulating the ATM kinase activity in the nucleus, at least partially.

RTS, a Rare Disease Associated with Moderate Radiosensitivity
Our findings suggest that fibroblasts derived from the RTS patients had cellular radiosensitivity, an abnormally high number of radiation-induced micronuclei, and a reduced number of early γH2AX and pATM foci, suggesting a delayed RIANS accompanied by the formation of cytoplasmic ATM-RECQL4 complexes and abnormal kinetics of MRE11 foci. The ZOPRA treatment permitted us to correct, at least partially, the radiosensitivity phenotype observed in the RTS fibroblasts. Since a subset of RTS patients (RTS2 subgroup) are generally at a high risk of cancer, some RTS patients may be exposed to anti-cancer radiotherapy ( Table 2 and references therein) but also to chemotherapy drugs, mimicking IR. Furthermore, in order to investigate some clinical features, RTS patients may also be exposed to repeated radiodiagnosis exams. Hence, such irradiating or DNAbreaking acts may raise the question of the potential worsening of some clinical symptoms of RTS patients accelerated by IR or chemotherapy drugs that provide DSB. The radiobiological characterization of the RTS and the role of the RECQL4 protein in the RI DNA damage repair and signaling therefore remain of both medical and scientific interest.
The SF2, the yields of micronuclei, and the number of γH2AX and pATM foci assessed post-irradiation are current biomarkers of radiosensitivity. They are quantitatively linked together [25,34,45]. Particularly, in a recent study, we provided inter-correlations between these biomarkers from a collection of 200 fibroblast cell lines eliciting a large spectrum of radiosensitivity [45]. The data described in this study were in very good agreement with each of the published correlations, therefore supporting their quantitative relevance. As indicated in Table 2, this study is the first work suggesting that skin fibroblasts from RTS patients are radiosensitive, impaired in the DSB recognition step and characterized by a delayed RIANS (Table 2). Interestingly, the cellular radiosensitivity quantified by the cell survival assay in the RTS fibroblasts was in quantitative agreement

RTS, a Rare Disease Associated with Moderate Radiosensitivity
Our findings suggest that fibroblasts derived from the RTS patients had cellular radiosensitivity, an abnormally high number of radiation-induced micronuclei, and a reduced number of early γH2AX and pATM foci, suggesting a delayed RIANS accompanied by the formation of cytoplasmic ATM-RECQL4 complexes and abnormal kinetics of MRE11 foci. The ZOPRA treatment permitted us to correct, at least partially, the radiosensitivity phenotype observed in the RTS fibroblasts. Since a subset of RTS patients (RTS2 subgroup) are generally at a high risk of cancer, some RTS patients may be exposed to anti-cancer radiotherapy ( Table 2 and references therein) but also to chemotherapy drugs, mimicking IR. Furthermore, in order to investigate some clinical features, RTS patients may also be exposed to repeated radiodiagnosis exams. Hence, such irradiating or DNA-breaking acts may raise the question of the potential worsening of some clinical symptoms of RTS patients accelerated by IR or chemotherapy drugs that provide DSB. The radiobiological characterization of the RTS and the role of the RECQL4 protein in the RI DNA damage repair and signaling therefore remain of both medical and scientific interest.
The SF2, the yields of micronuclei, and the number of γH2AX and pATM foci assessed post-irradiation are current biomarkers of radiosensitivity. They are quantitatively linked together [25,34,45]. Particularly, in a recent study, we provided inter-correlations between these biomarkers from a collection of 200 fibroblast cell lines eliciting a large spectrum of radiosensitivity [45]. The data described in this study were in very good agreement with each of the published correlations, therefore supporting their quantitative relevance. As indicated in Table 2, this study is the first work suggesting that skin fibroblasts from RTS patients are radiosensitive, impaired in the DSB recognition step and characterized by a delayed RIANS (Table 2). Interestingly, the cellular radiosensitivity quantified by the cell survival assay in the RTS fibroblasts was in quantitative agreement with the literature data. Particularly, the radiosensitivity of the RTS fibroblasts was similar to that found in cells from Bloom, Werner, Cockayne, and some xeroderma pigmentosum syndromes, i.e., genetic syndromes caused by mutated helicase [45,59]. However, despite such coherence, since RTS is a very rare syndrome and since few untransformed fibroblast cell lines derived from RTS patients are available, further investigations are needed to generalize our observations to other RTS cases.

The Radiosensitivity Associated with the RTS and the RIANS Model
How can we explain the level of radiosensitivity observed in the RTS cells? By considering the helicase function of the RECQL4 protein, it appeared natural to directly explain the clinical features specific to the RTS by impaired DNA helicase activity. Besides, the RECQL4 protein has been found to be involved in the choice between the NHEJ and the HR and in the signaling and repair of DSB, the key-damage induced by RI [11,[20][21][22][23][24]60]. However, the fibroblasts used in the present study were quiescent. Hence, the impact of the HR on the RI lethal effect was likely to be negligible. Secondly, the cellular models used in the pre-cited studies are tumors that were transformed by sophisticated vectors to over-express the wild-type or the mutated RECQL4 gene [11,[20][21][22][23][24]60]. However, even if an impaired helicase (whose function can be redundant) may explain the significant but moderate radiosensitivity observed in RTS cells, it must be stressed that the cytoplasmic expression of RECQL4 was systematically higher in the RECQL4-mutated fibroblasts than in the radioresistant controls. Lastly, no combination of statins and bisphosphonates was shown to influence DNA helicase activity while the ZOPRA treatment may complement the radiosensitivity phenotype of the three RTS cell lines tested, at least partially: the radiosensitivity associated with the RTS suggests a more complex model.
In the framework of the RIANS model, the oxidative stress produced by IR triggers the formation of DNA damage in the nucleus (notably the DSB) and the monomerization of the ATM dimers (more abundant in the cytoplasm). The ATM monomers diffuse in the nucleus and trigger the recognition of DSB by the NHEJ pathway via the phosphorylation of H2AX by ATM. This scenario corresponds to the radioresistance (so-called group I, in the frame of RIANS model). In the case of delayed RIANS (group II, in the framework of the RIANS model), the ATM monomers are sequestrated in the cytoplasm by over-expressed ATM substrates, called X-proteins. Group II cells show significant but moderate radiosensitivity and are more susceptible to transformation, similar to the RTS cells. Conversely, any acceleration of the RIANS, such as that induced by ZOPRA, may render cells more radioresistant [43]. The X-proteins should therefore reach two major requirements: (1) be overexpressed in the cytoplasm; (2) associate with and/or be phosphorylated by ATM. While we observed abundant cytoplasmic forms of RECQL4 in the three RTS fibroblast cell lines tested here, can the RECQL4 protein be an X-protein in the framework of the RIANS model? Kim et al. have shown that the ATM kinase preferentially phosphorylates the SQ/TQ domains [61]. Interestingly, the RECQL4 protein has five SQ domains (including Ser27 situated in the SLD2 domain and Ser217, Ser326, and Ser 611 situated in the DEAD domain of the core helicase domain) [62]. The PLA data clearly suggested that ATM and RECQL4 preferentially interact in the cytoplasm, which is likely to cause a delay in the RIANS and therefore an alteration of the DSB recognition by the NHEJ pathway ( Figure 10). Hence, by considering both the impaired helicase activity but also the role of ATM substrates as X-proteins, mutations of the RECQL4 protein may be linked to an abnormal response to IR. However, further investigations are needed to identify the molecular and cellular role of the phosphorylation of RECQL4 by ATM in quiescent normal and RTS fibroblasts. Figure 10. The RTS and the RIANS model. Schematic illustration of the RIANS model for the radioresistant and the RTS cells. The IR-induced ATM monomers diffuse in the nucleus, in which they trigger the DSB repair pathway by phosphorylating H2AX (activation of NHEJ) and MRE11 (inhibition of the nuclease activity). In the RTS cells, RECQL4 mutations lead to impaired cytoplasmic localization and/or overexpression of the RECQL4 protein, which results in more ATM-RECQL4 complexes: a significant delay of RIANS occurs with a lack of DSB recognition and repair and increased MRE11 activity. This delay of RIANS may be alleviated by ZOPRA treatment.

Conclusions
Our findings indicate that the radiosensitivity and the sensitivity to DNA breaking agents associated with RTS may be caused by an overexpression of some cytoplasmic RECQL4 helicase forms that sequester ATM monomers in the cytoplasm, which leads to a delayed RIANS, impaired DSB recognition, and high genomic instability. Further experiments are required to better quantify the risks caused by radiodiagnosis, radiotherapy or chemotherapy in the healthy tissues of RTS patients and to investigate whether statins, bisphosphonates, and/or anti-oxidative drugs may provide efficient radioprotection.  Institutional Review Board Statement: As already mentioned in the Materials and Methods, this study was performed with human fibroblast cell lines that belong to the COPERNIC collection of the laboratory that has been approved by the regional Ethical Committee. Cell lines were declared under the numbers DC2008-585, DC2011-1437, and DC2021-3957 to the Ministry of Research. Figure 10. The RTS and the RIANS model. Schematic illustration of the RIANS model for the radioresistant and the RTS cells. The IR-induced ATM monomers diffuse in the nucleus, in which they trigger the DSB repair pathway by phosphorylating H2AX (activation of NHEJ) and MRE11 (inhibition of the nuclease activity). In the RTS cells, RECQL4 mutations lead to impaired cytoplasmic localization and/or overexpression of the RECQL4 protein, which results in more ATM-RECQL4 complexes: a significant delay of RIANS occurs with a lack of DSB recognition and repair and increased MRE11 activity. This delay of RIANS may be alleviated by ZOPRA treatment.

Conclusions
Our findings indicate that the radiosensitivity and the sensitivity to DNA breaking agents associated with RTS may be caused by an overexpression of some cytoplasmic RECQL4 helicase forms that sequester ATM monomers in the cytoplasm, which leads to a delayed RIANS, impaired DSB recognition, and high genomic instability. Further experiments are required to better quantify the risks caused by radiodiagnosis, radiotherapy or chemotherapy in the healthy tissues of RTS patients and to investigate whether statins, bisphosphonates, and/or anti-oxidative drugs may provide efficient radioprotection. Informed Consent Statement: All anonymous donors of the COPERNIC collection mentioned above were informed and signed consent according to the ethics and regulatory recommendations, notably for their consent to participate and for publication of anonymous data.