The human EA.hy926 endothelial cells were obtained from the American Type Culture Collection (ATCC; Manassas, VA, USA). They were cultured (37°C-5% CO₂) in Dulbecco’s modified Eagle’s medium supplemented with 10% fetal bovine serum and 1% penicillin/streptomycin (all from N.V. Invitrogen S.A., Merelbeke, Belgium). The cells were regularly examined for the absence of mycoplasma using the LookOut^® Mycoplasma PCR Detection kit (Sigma-Aldrich, St. Louis, MO, USA).

The cells were seeded at a density of 10⁵ cells in 12.5 cm² flasks. Twenty-four hours after plating, the flasks were placed in a transportable incubator (37°C) and moved from the resident laboratory (Mol, Belgium) to the GSI Helmholtzzentrum für Schwerionenforschung GmbH (Darmstadt, Germany). Forty-eight hours after plating, the subconfluent cells were irradiated in flasks completely filled with culture medium with a 1 GeV/u Ni beam at the SIS facility at GSI with the intensity controlled raster scanning technique as described by Haberer et al (20). The ion energy at the sample position was approximately 930 MeV/u with a LET of 183 keV/μm (calculated with the program code ATIMA). The culture flasks were placed vertically and exposed perpendicularly to the nickel ion beam at the following doses: 0.5, 2 and 5 Gy. Non-irradiated control samples were treated similarly to the irradiated samples, but placed out of the beam. Following irradiation, the cells were incubated (37°C, 5% CO₂) in 2 ml of conditioned medium until fixation time points (2, 8 and 24 h).

The cells were fixed in 4% paraformaldehyde (Merck KGaA, Darmstadt, Germany) 2 and 24 h after irradiation. They were then treated with 0.25% Triton X for 5 min, blocked with 3% bovine serum albumin (both from Sigma-Aldrich) for 30 min and incubated overnight with mouse anti-γ-H2AX antibody (Abcam, Cambridge, MA, USA) at 4°C. After a second blocking of 10 min, the cells were incubated for 1 h with anti-mouse secondary antibody coupled to FITC (Sigma-Aldrich) at 37°C and then mounted in Vectashield mounting medium (Vector Laboratories, Burlingame, CA, USA) with DAPI. Between each of the previous steps, the slides were washed with phosphate-buffered saline (PBS).

An automated inverted fluorescence microscope (TE2000-E; Nikon, Tokyo, Japan), equipped with a motorized XYZ stage, emission and excitation filter wheels, shutters and a triple dichroic mirror (436/514/604) was used for the image acquisition of the immunostained slides. Images were acquired with a 40X Plan Fluor oil objective (NA 1.3) and an Andor iXon EMCCD camera (Andor Technology, South Windsor, CT, USA). For each sample, at least 12 fields were acquired on 5 z-stack focusses (1 μm). The γ-H2AX spot number and spot occupancy were analyzed with the INSCYDE plugin for ImageJ as previously described (21). Spot occupancy was defined for each nucleus as the sum of the spot areas divided by the nucleus area (spot_occupancy = sum (spot_area)/nuclear_area). A minimum number of 100 cells was analyzed in 2 biological replicates per condition. For statistical analyses, the data were analyzed using the Mann-Whitney U test with SPSS version 17.0 software (IBM Corp., Chicago, IL, USA) and box plots were generated using the same software. P-values <0.05 were considered to indicate statistically significant differences.

At 2 time points after irradiation (8 and 24 h), the adherent cells were washed in PBS, lysed in 350 ml of AllPrep DNA/RNA/Protein Mini kit lysis buffer (Qiagen, Hilden, Germany) and frozen at −80°C. RNA was extracted using the same kit and its concentration was measured using a NanoDrop spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA), while its quality (RNA integrity number, RIN) was determined using Agilent’s lab-on-chip Bioanalyzer 2100 (Agilent Technologies, Santa Clara, CA, USA). All RNA samples had a RIN value >9.0.

RNA was processed using the GeneChip WT cDNA Synthesis and Amplification kit (Affymetrix, Santa Clara, CA, USA) according to the manufacturer’s instructions. The resulting RNA was hybridized to Affymetrix Human Gene 1.0 ST arrays which contain an estimated number of 28,869 genes based on the March 2006 [UCSC Hg 18; National Center for Biotechnology Information (NCBI) build 36] human genome assembly. Biological triplicates were collected for each condition.

Raw data (.cel-files) were imported at exon level in Partek Genomics Suite version 6.5 (Partek, Inc., St. Louis, MO, USA). Briefly, robust multi-array average (RMA) background correction was applied, data were normalized by quantile normalization and probeset summarization was performed by the median polish method. Gene summarization was performed using one-step Tukey’s biweight method. The obtained data were analyzed with Partek Genomics Suite for single gene analysis. One- or two-way ANOVA, taking into consideration the scan date (where applicable) and the dose as factors, were performed for each time point. In order to determine statistical significance, thresholds were set on the p-value <0.001 and on the fold-change >1.5.

The enrichment of the transcription factor binding motifs was analyzed using Pscan Ver. 1.2 (22) and the JASPAR database, scanning in a region from −950 to +50 base pairs from the transcription start site.

To assess DNA damage induction by nickel ion irradiation and evaluate the cell capacity required to repair this damage, we performed a high content cytometric assay of γ-H2AX, 2 and 24 h after exposure. As measured by the number of γ-H2AX foci, DNA damage was significantly increased 2 h after nickel ion irradiation (2 Gy), with an average number of 15 foci per nucleus vs. 2 foci per nucleus in the control samples (Fig. 1A). Twenty-four hours after irradiation, the number of foci decreased to 9 per nucleus, which was significantly higher than the values of controls, indicating that part of the DNA damage persisted for at least 24 h. Similar trends were observed for the spot occupancy, which is the fraction of the projected area of the nucleus occupied by the signal from the γ-H2AX foci (Fig. 1B).

In order to evaluate gene expression, we performed microarrays 8 and 24 h after irradiation. A 0.5 Gy irradiation, both after 8 and 24 h, elicited a subtle effect on gene expression in the EA.hy926 cells. Six annotated genes were differentially regulated with fold changes (FC) between 1.5 and 1.8 after 8 h, and 18 genes were differentially regulated with an FC between 1.5 and 2.3 after 24 h. A more drastic effect was observed at 5 Gy, 24 h after irradiation. At this time point, we detected the upregulation of 77 annotated genes (Fig. 2 and Table I; maximum FC, 3.4). Among these genes, cytokines and chemokines (CXCL5, TGFA, TRIM22, TNFSF9, EBI3, IL-6, IL-11 and CD70) were identified, as well as genes involved in DNA damage response (SPATA18, POLL, APOBEC3H and SESN1), cell cycle arrest (ZMAT3, MXD4, TP53INP1, HSPB8, TGFA, SESN2, BTG2, DTX3, TOB1, HBP1, CDKN1A and PLK3) and apoptosis (TP53INP1, HSPB8, TGFA, TP53I3, MOAP1, CYFIP2, TRADD, DTX3 and FAS). In addition, we observed the upregulation of genes coding for ion channels (SLC22A4, KCNJ2, ORAI3 and CLIC3), cell adhesion (CEACAM1 and NEU1) and oxidative stress response proteins (FMO4, FDXR, SIRT2 and SESN1).

A total of 145 annotated genes was downregulated 24 h after nickel ion irradiation (5 Gy) (Fig. 3 and Table I). The majority (62 genes) is known to be involved in various aspects of cell division, such as DNA replication, replication forks and chromosome assembly and segregation (Table II and Fig. 4). Other downregulated genes found have been implicated in post-replicative DNA repair (UNG, UPF3A, MSH2 and MSH6), nucleotide biosynthesis (DHFR and RRM2), DNA repair (FANCA, MMS22L, NFKBIL2, RAD51, EXO1 and HMGB2), positive (YAP1) and negative regulation of apoptosis (DHRS2, DHCR24 and MTBP), Rho signaling (ARHGAP19, ARHGAP11B and RACGAP1) and cell adhesion (PVRL1 and DLGAP5).

In order to identify the transcription factors potentially responsible for the differential gene expression upon irradiation, we scanned sequences close to the transcription start sites of these genes using Pscan (22). We found motifs for E2F1 among the transcription factor binding motifs enriched in the downregulated gene list, (p-value <10–19). On the other hand, we found two members of the REL family (RelA and NF-κB) with significantly enriched binding motifs in the list of upregulated genes (p-values <0.05).

We measured a significant increase in the number of γ-H2AX foci 2 h following nickel ion irradiation. This number was lower than the 30 spots per nucleus that we measured on average upon X-irradiation with the same dose (data not shown). However, it is not so surprising since high LET irradiation deposits high amounts of energy along well-separated tracks. For nickel ions with a LET of 183 keV/μm and at a dose of 2 Gy, we calculated an average of 6.8 direct particle hits per nucleus (100 μm²), which follows a Poisson distribution. However, we observed an average of 15 spots per nucleus. This may be due to the secondary radiation from the ion track and the basal level of endogenous γ-H2AX foci as observed in the controls.

Considering that the imaging of γ-H2AX foci was performed at the same angle as ion tracks produced by the irradiation beam, the complexity of the damage along these tracks could not be taken into account. However, the DNA damage complexity is known to be important in high LET irradiation (23–26). Although significantly increased, the γ-H2AX spot occupancy did not seem to be able to account for the complexity of DNA damage and showed similar results to the spot number measurement. This complex DNA damage is associated with slower repair (27) and therefore leads to a more pronounced delayed cellular damage (26). Our results revealed a significant level of γ-H2AX foci 24 h following nickel ion irradiation, as compared to controls; this suggests the presence of complex DNA damage.

Nickel ion irradiation at a dose of 0.5 Gy elicited a lower gene expression response as compared to a dose of 5 Gy, in terms of the number of regulated genes and FC. At 24 h post-irradiation (5 Gy), we observed an upregulation of 12 genes involved in cell cycle arrest and a downregulation of 62 genes involved in cell cycle progression, among which were 3 members of the E2F transcription factor family (E2F1, E2F2 and E2F8). Moreover, the transcription factor binding motifs for E2F1 were found to be highly enriched in the list of downregulated genes. E2F is a family of transcription factors known to control G1- to S-phase transition (28), and to regulate the expression of a large variety of genes involved in DNA replication, DNA repair and apoptosis (29). Among the E2F transcription factors, E2F1 is known to be stabilized upon DNA damage through its phosphorylation by ataxia telangiectasia-mutated (ATM) kinase, ATM and Rad3-related (ATR) kinase and checkpoint kinase 2 (CHK2), as well as through its acetylation (29). Our results suggest a major role of E2F transcription factors in the response of EA.hy926 cells to high LET irradiation.

Six components of the minichromosome maintenance (MCM) complex, a heterohexamer helicase essential for the initiation and elongation step of DNA replication (30), were downregulated. This helicase may be a target for replication checkpoints (31), and is thought to be regulated mostly through post-transcriptional modifications (32). However, our results indicate a possible transcriptional regulation of several members of the MCM complex. Apart from MCM, many of the observed downregulated genes are involved in DNA replication and in chromosome formation, maintenance and segregation, indicating their key role in cell cycle regulation in response to high LET radiation. Of note, we also reported the downregulation of 4 genes involved in post-replication DNA repair (UNG, UPF3A, MSH2 and MSH6), which may be silenced in the absence of active replication.

During this study, irradiation was performed on proliferating endothelial cells. The results gathered on cell cycle gene expression are therefore of moderate interest for mature blood vessels where proliferation is limited. However, as far as hadron therapy is concerned, our data indicate that high LET radiation may have a significant impact on the cellular proliferation of newly formed vascular vessels in the vicinity of the targeted tumor.

The expression of several genes involved in the DNA damage response, oxidative stress response and apoptosis was induced 24 h after 5 Gy of nickel ion irradiation, with a concomitant reduction of genes involved in DNA repair. However, these effects were not significant at a dose of 0.5 Gy, at either time points (8 and 24 h). These results suggest that a high dose of nickel ion irradiation induces a global DNA damage response, accompanied by cell cycle arrest and an increase in pro-apoptotic gene expression 24 h after irradiation.

The impermeability of the endothelium is essential for the vasculature integrity and is determined by the cooperation of cell junctions and the cytoskeleton (33,34). In turn, adhesion molecules regulate cell homeostasis, growth and apoptosis (33). A number of cellular pathways are known to regulate cell adhesion in endothelial cells. These include growth factors, Rho GTPases, protein kinases and calcium signaling (34,35). The alteration of these pathways or of adhesion molecules may trigger the radiation-induced retraction observed by others in endothelial cells (13,14). Our study identified the differential expression of a number of genes known to be involved in cell adhesion (CEACAM1 and NEU1), cytoskeleton architecture (TUBA4A, LIMA1 and PLS1), Rho signaling (ARHGAP19, ARHGAP11B and RACGAP1) and calcium metabolism (ORAI3, CAMK2N1 and CALML4) 24 h after 5 Gy of nickel ion irradiation, which are potentially involved in endothelial cell retraction.

Inflammatory responses mediated by endothelial cells are believed to be involved in radiation-induced cardiovascular disease (7). Our study revealed the upregulation of 8 cytokines or chemokines that may be linked to inflammation (CXCL5, TGFA, TRIM22, TNFSF9, EBI3, IL-6, IL-11 and CD70). Of note, a search for transcription factor binding motifs that are significantly enriched in the list of upregulated genes upon 5 Gy of irradiation, revealed 2 members of the REL family (RelA and NF-κB). This family of transcription factors induces the expression of a multitude of genes, such as cytokines, proliferation, pro-survival and anti-apoptotic genes (36). For instance, we found IL-6 to be upregulated after 5 Gy of nickel ion irradiation. IL-6 expression was also shown to be upregulated by low-dose radiation therapy (10). IL-6 is known to be activated by NF-κB (36,37) and is thought to play a role in radiation-induced cardiovascular disease (1,7). The secretion of cytokines may also affect non-irradiated cells by a bystander effect. Indeed, in human fibroblasts, the external addition of IL-6 has been shown to increase γH2AX spot occupancy (38). The activation of NF-κB may be linked to the transcription factor, signal transducer and activator of transcription 3 (STAT3) (37), of which we also found significant binding motif enrichment.

In conclusion, we observed a downregulation of multiple genes involved in cell division, particularly at 24 h after nickel ion irradiation. Our results suggest an important role for E2F transcription factors in this process. The endothelial function being based on a plethora of intercellular interactions within a dynamic structure involving cell movements and turnover, cell cycle arrest may play a role in the radiation-induced cardiovascular disease. On the other hand, we observed an upregulation of various cytokines which may be induced by NF-κB. Other studies have also suggested that these cytokines may be linked to radiation-induced cardiovascular disease (10). The effects on gene expression were observed upon high doses of acute irradiation and are less relevant to space exploration. However, during hadron therapy, healthy tissues surrounding tumors, such as endothelial cells, may be subjected to high doses, which may lead to complications. In this study, we identified a multitude of potential molecular targets for further mechanistic studies out of which the gene expression changes upon high doses of nickel ion irradiation may be important for patients treated with hadron-therapy.