The authors have declared that no competing interests exist.
Conceived and designed the experiments: AW MHR MJA KFB SH ST. Performed the experiments: RY OA JM. Analyzed the data: RY OA AS JM ST. Contributed reagents/materials/analysis tools: AS KM JM KFB. Wrote the paper: RY OA MJA SH ST.
The etiology of radiation-induced cardiovascular disease (CVD) after chronic exposure to low doses of ionizing radiation is only marginally understood. We have previously shown that a chronic low-dose rate exposure (4.1 mGy/h) causes human umbilical vein endothelial cells (HUVECs) to prematurely senesce. We now show that a dose rate of 2.4 mGy/h is also able to trigger premature senescence in HUVECs, primarily indicated by a loss of growth potential and the appearance of the senescence-associated markers ß-galactosidase (SA-ß-gal) and p21. In contrast, a lower dose rate of 1.4 mGy/h was not sufficient to inhibit cellular growth or increase SA-ß-gal-staining despite an increased expression of p21. We used reverse phase protein arrays and triplex Isotope Coded Protein Labeling with LC-ESI-MS/MS to study the proteomic changes associated with chronic radiation-induced senescence. Both technologies identified inactivation of the PI3K/Akt/mTOR pathway accompanying premature senescence. In addition, expression of proteins involved in cytoskeletal structure and EIF2 signaling was reduced. Age-related diseases such as CVD have been previously associated with increased endothelial cell senescence. We postulate that a similar endothelial aging may contribute to the increased rate of CVD seen in populations chronically exposed to low-dose-rate radiation.
Cardiovascular disease (CVD) – pathologies of the heart, blood vessels and the vascular system of the brain– is the leading cause of morbidity and mortality in the Western world
A large group of people receive repetitive low-dose exposures from medical radiation for imaging purposes
The vascular endothelium is a strong target candidate for radiation due to its intrinsic high level of sensitivity to radiation
At sites overlying atherosclerotic plaques, an increase in senescent endothelial cells have been found in the human aorta and coronary arteries, indicating that endothelial senescence is associated with normal plaque formation in non-exposed populations
A study using a subset of animals from the Janus series of experiments compared acute or fractionated exposures of gamma or neutron radiation on the hazards associated with the development of cancer and non-cancer diseases of the liver, lung, kidney or vascular system
We have shown previously that chronic low-dose-rate (4.1 mGy/h) exposure is able to trigger premature senescence in HUVECs. Our data suggested that chronic radiation-induced DNA damage and oxidative stress activate the p53/p21 pathway
The aim of the present study was to investigate whether continuous radiation exposure using even lower dose rates (1.4 mGy/h and 2.4 mGy/h) than in the previous study could induce premature senescence in HUVECs. The effect of these dose rates on biological pathways associated with normal senescence was investigated using reverse phase protein arrays (RPPA) and Isotope Coded Protein Label (ICPL) technology in combination with tandem mass spectrometry (LC-MS/MS). This study raises the question whether there is a lower limit in the dose rate at which there is no measurable induction of premature senescence.
Ionizing radiation was performed in a custom-made cell culture incubator modified to hold a 137Cs-gamma source exposing cells to dose rates of 1.4 mGy/h and 2.4 mGy/h. Irradiation was carried out continuously except during replacement of culture medium and passaging of cells which lasted from 30 min to 1 hour. Sham irradiated cells (control) were grown in an identical incubator but without exposure to radiation.
HUVECs (Invitrogen, Paisley, UK) were obtained from a single donor and cultured in Media 200 (Invitrogen, Paisley, UK) supplemented with low serum growth supplement (LSGS) containing 2% fetal bovine serum, 1 µg/µl hydrocortisone, 10 ng/ml epidermal growth factor, 10 µg/ml heparin, 100 U/ml penicillin and 0.1 mg/ml streptomycin at 37°C in a 95% air/5% CO2 humidified atmosphere. Cells were passaged every seven days (5000 cells/cm2) with the culture medium changed every two days. Cells were passaged using accutase (Invitrogen, Paisley, UK) and growth rate kinetics calculated using the equation:
Statistical significance of the slopes in the growth curves were calculated using a polynomial second order equation (Y = A+B*X+C*Xˆ2) with the software Graph Pad Prism.
SA-ß-gal activity was determined using a histochemical staining kit according to the manufactureŕs instructions (Sigma-Aldrich). Briefly, HUVEC monolayers were rinsed twice in phosphate buffered saline (PBS), fixed at room temperature for 6–7 of the slopes in the growth curves were calculated min in 2% formaldehyde/0.2% glutaraldehyde, washed three times in PBS and finally incubated at 37°C with SA-ß-gal staining solution (1 mg/ml 5-bromo-4-chloro-3-indolyl ß-D galactoside, Sigma-Aldrich, Sweden) in buffer containing 100 mM citric acid, 200 mM sodium phosphate, 5 mM potassium ferrocyanide, 5 mM potassium ferricyanide, 150 mM NaCl, and 2 of the slopes in the growth curves were calculated mM MgCl2 at pH 6.0. The percentage of SA-ß-gal stained HUVECs was determined by counting at least 1,000 cells using light microscopy.
The harvested HUVECs from week 1, 6 and 10 were washed twice in ice cold phosphate buffered saline and centrifuged at 4°C, 200 g for 10 min. Protein extraction was performed by dissolving the cell pellet in extraction buffer EXBplus (Qiagen, Hilden Germany) and incubated on wet ice for 15 minutes followed by incubation at 95°C for 10 min. Cells were centrifuged at 4°C, 20,000g for 15 min. Supernatants were collected and stored at −80°C until further analysis. Protein concentrations were determined using the Bradford protein assay (Bio-Rad laboratories, Munich, Germany).
RPPA analysis was performed as previously described
# | Primary antibody (PAb) | Type | 1.4 mGy/h week 10 | 2.4 mGy/h week 10 | Mw (kDa) | Company name | Number | Source |
1 | Akt | Signal transduction | + | Un | 60 | Cell signalling | #9272 | Rabbit |
2 | Caldesmon | Cytoskeletal protein | Un | – | 70–80, 120–150 | Cell signalling | #2980 | Rabbit |
3 | Caspase 3 | Apoptosis | Un | Un | 17, 19, 35 | Cell signalling | #9662 | Rabbit |
4 | Desmin | Cytoskeletal protein | Un | Un | 53 | Cell signalling | #5332 | Rabbit |
5 | Fascin | Cytoskeletal protein | Un | Un | 55 | Abcam | ab78487 | Mouse |
6 | HIF1-alpha | Transcription factor | Un | Un | 120 | BD Biosciences | 610959 | Mouse |
7 | Hsp 60 | Stress responsive proteins | Un | Un | 60 | Abcam | #ab46798 | Rabbit |
8 | mTOR | Serine/threonine protein kinase | – | – | 298 | Cell signalling | #4517 | Rabbit |
9 | NF-kB p65 | Transcription factor | Un | Un | 65 | Cell signalling | #3034 | Mouse |
10 | Nucleolin | Nucleolar protein | – | Un | 100 | EMD Millipore | 05–565 | Mouse |
11 | p38 MAPK | Signal transduction | Un | Un | 43 | Cell signalling | #9212 | Rabbit |
12 | p44/42 MAPK | Signal transduction | Un | Un | 42, 44 | Cell signalling | #9102 | Rabbit |
13 | Phospho NF-kB p65 (Ser536) | Transcription factor | Un | Un | 75 | Cell signalling | #3031 | Rabbit |
14 | Phospho PTEN (Ser380) | Tumor suppressor | Un | Un | 54 | Cell signalling | #9551 | Rabbit |
15 | Phospho-Akt (Ser473) | Signal transduction | – | – | 60 | Cell signalling | #4060 | Rabbit |
16 | Phospho-HSP27 (Ser 78) | Stress responsive proteins | Un | Un | 27 | Cell signalling | #2405 | Rabbit |
17 | Phospho-p38 MAPK (Thr108/Tyr182) | Signal transduction | Un | Un | 43 | Cell signalling | #4631 | Rabbit |
18 | Phospho-p44/42 MAPK (Thr202/Tyr204) | Signal transduction | – | – | 42, 44 | Cell signalling | #9101 | Rabbit |
19 | Phospho-STAT3 (Tyr705) | Transcription factor | Un | Un | 79,86 | Cell signalling | #9145 | Rabbit |
20 | Phospho-STAT5 (Tyr694) | Transcription factor | Un | Un | 90 | Cell signalling | #9358 | Rabbit |
21 | Phospho-VEGFR2 | Platelet-derived growth factor | Un | Un | 230 | Cell signalling | #2478 | Rabbit |
22 | PI3K | Intracellular signal transducer | – | – | 85 | Cell signalling | #4292 | Rabbit |
23 | PTEN | Tumor suppressor | Un | Un | 54 | Cell signalling | #9552 | Rabbit |
24 | Ras | Ras family | Un | Un | 21 | Cell signalling | #3965 | Rabbit |
25 | Rho GDI | Rho GDI pathway | – | – | 26 | Epitomics | 2751–1 | Rabbit |
26 | ROCK | Rho-associated Kinase | Un | Un | 160 | Cell signalling | #4035 | Rabbit |
27 | SAPK/JNK | Signal transduction | Un | Un | 46,54 | Cell signalling | #9252 | Rabbit |
28 | SAPK/JNK-Phospho (Thr183/Tyr185) | Signal transduction | Un | Un | 46, 54 | Cell signalling | #9255 | Mouse |
29 | Serpine (PAI1) | Plasminogen activator inhibitor | Un | Un | 45 | Serotec | A8P1100 | Goat |
30 | Snail | Transcription factor | Un | Un | 37 | Selber-gemacht | Rat | |
31 | STAT3 | Transcription factor | Un | Un | 79,86 | Cell signalling | #9132 | Rabbit |
32 | STAT5 | Transcription factor | Un | Un | 90 | Cell signalling | #9359 | Rabbit |
33 | SUMO1 | Small ubiquitin-like modifier | Un | Un | 90 | Santa cruz | Sc5308 | Mouse |
34 | VEGFR-2 | Platelet-derived growth factor | Un | Un | 210, 230 | Cell signalling | 55B11 | Rabbit |
35 | Vimentin | Cytoskeletal protein | Un | Un | 57 | DAKO | M0725 | Mouse |
The antibodies were chosen to represent senescence-associated proteins. Differentially regulated proteins at week 10 are marked with + = up-regulated, − = down-regulated or “un” = unregulated.
To assess protein expression quantitatively TIF files of the Sypro Ruby and antibody stained slides were generated using a Scanjet 3770 gray scale scanner (Hewlett-Packard, Hamburg, Germany). The intensity of data points was quantified utilizing MicroVigene 3.5.0.0 software (Vigenetech, Carlisle, MA). The MicroVigene signal-intensity (MVS) was calculated as the integral of a logistic four-point fit model, which was matched optimally to the 12 data time points that were obtained. The MVS for each antibody was normalized to total protein MVS. Results were evaluated by t-test and a value of p≤0.05 was considered to denote statistical significance. Data are reported as ± SEM.
Proteins were isolated from exposed and control cells after washing with PBS using lysis buffer. Proteins were then separated by SDS-PAGE and transferred to nitrocellulose membranes (GE Healthcare) using a TE 77 semi dry blotting system (GE Healthcare) at 1 mA/cm for 1 h. Nitrocellulose membranes were blocked using 3 % BSA in PBS, pH 7.4, for 1 h at room temperature, washed three times in 10 mM Tris-HCl, pH 7.4, with 150 mM NaCl for 5 min and incubated over night at 4°C with antibodies against p21 (#2947, Rabbit monoclonal, Cell Signaling), p53 (Sc126, Mouse monoclonal, Santa Cruz Biotechnology), and phospho-p53 (Ser-15) (#9286, Mouse monoclonal, Cell Signaling), p16 (Sc-81613, Santa Cruz Biotechnology) using dilutions recommended by the manufacturer. Actin (sc1616, Santa Cruz Biotechnology) was used as a loading control. After washing three times, blots were incubated with the appropriate horseradish peroxidase-conjugated secondary antibodies (Santa Cruz Biotechnology) for 1 h at room temperature and developed using the ECL system (GE Healthcare) following standard procedures. Quantification of immunoblot bands was performed on digitized images using TotalLab TL100 software (
Passaged cells at culture week 10 from sham, 1.4 mGy/h and 2.4 mGy/h dose rates were collected for protein analysis using accutase (Invitrogen, Paisley, UK) for detachment. All experiments were performed with three biological replicates. Proteins were isolated using Isotope Coded Protein Labeling (ICPL) lysis buffer (SERVA Electrophoresis GmbH, Germany) and the protein concentration determined by Bradford assay as above. The proteins were extracted form cells cultured for 10 weeks and labeled with ICPL triplex reagent (SERVA) according to the manufacturer's protocol. Schematic representation of labeling and analysis parameters from three replicates is shown in
The protein samples from week 10 (control and two irradiated) were reduced and alkylated before labeling with ICPL0, ICPl4 and ICPL6. Samples were mixed and further separated using 1D gel electrophoresis and digested as described in Methods. Samples were analyzed by LC-ESI-MS/MS. Quantification of proteins was performed by Proteome Discoverer software using three biological replicates.
After Coomassie blue staining, each lane in the gel was cut into 5 equal slices and subjected to in-gel digestion with trypsin (Sigma Aldrich) as described previously
Eluted peptides were analyzed by a LTQ Orbitrap XL mass spectrometer (Thermo Scientific) equipped with a nanospray ionization source, operated in the data-dependent mode to automatically switch between Orbitrap-MS and LTQ-MS/MS acquisition. Full scan MS spectra (from m/z 300 to 1500) were acquired in the Orbitrap with a resolution of 60,000. Up to ten of the most intense ions were selected for fragmentation in the linear ion trap using collision-induced dissociation, depending on signal intensity. High resolution MS scans in the Orbitrap and MS/MS scans in the linear ion trap were acquired in parallel. Target peptides already selected for MS/MS were dynamically excluded for 60 seconds.
The MS/MS spectra were searched against the ENSEMBL Human database (Version: 2.4, 96580 sequences) using the MASCOT search engine (version 2.3.02; Matrix Science) with the following parameters: a precursor mass error tolerance of 10 ppm and a fragment tolerance of 0.6 D. One missed cleavage was allowed. Carbamidomethylation of cysteine was set as a fixed modification. Oxidized methionine and ICPL 0, ICPL-4, and ICPL-6 for lysine were set as variable modifications.
Data processing for protein identification and quantification of ICPL- labeled proteins was performed using Proteome Discoverer version 1.3.0.339 (Thermo Scientific) as described before
The technical variation between replicates was validated using spiked protein test mixture containing bovine carbonic anhydrase with light to medium to heavy ratio of 1∶1∶2. Significant fold changes were determined by technical variability based on the average values of the CV (10.5%) of spiked protein. Proteins identified by at least 2 unique peptides in two out of three replicates, quantified by the H/L and M/L variability of 2–30% and with a p-value smaller than 0.05 (t-test) were considered for further evaluation. We considered the variability of 30% (>2 CV) as significant as it overcomes technical variability in our experiments. Therefore, proteins with ratios of H/L label greater than 1.3-fold or less than −1.3-fold were defined as significantly differentially expressed. The biological significance of this fold change cut off is in good agreement with the previously published data
Network analysis was performed from the proteomic data with the software tool Ingenuity Pathway Analysis (IPA) (INGENUITY System,
The number of cumulative population doublings (CPD) was measured for control and ±0.4 CPD (
A) The growth curve is plotted with cumulative population doublings versus time. Growth curves of control (blue), 1.4 mGy/h (orange) and 2.4 mGy/h (green) irradiated HUVECs are shown. Cumulative population doublings from each week are presented as means ± SEM (n = 3). B) Histograms of positively-stained cells for SA-ß-gal (senescence marker) for control (blue), irradiated by 1.4 mGy/h (orange) and irradiated by 2.4 mGy/h (green). Data are presented as means ± SEM (n = 3). (Students t-test; *p<0.05, **p<0.01 and ***p<0.005).
Statistical analysis with the polynomial second order equation showed no significant change between the slopes of sham irradiated and 1.4 mGy/h curves. Whereas the slope of 2.4 mGy/h curve was significantly different from both sham irradiated and 1.4 mGy/h.
The number of senescent cells was estimated by SA-ß-gal staining at weeks 1, 6, 10 and 12 (
Analysis of the changes in protein expression accompanying senescence was carried out using cells harvested at weeks 1, 6 and 10. The cumulative radiation doses at these time points were 0.40 Gy, 2.41 Gy and 4.03 Gy for the higher 2.4 mGy/h dose rate, and 0.24 Gy, 1.41 Gy and 2.35 Gy for the lower 1.4 mGy/h dose rate and respectively.
Cell cycle inhibitor protein p21, a known marker of senescence
A representative immunoblot of p21 expression levels at different time points and dose rates is shown (A). The columns represent protein levels of p21 (B) Akt (C), phospho-Akt (D), PI3K (E) and mTOR (F) in control (blue), 1.4 mGy/h (orange) and 2.4 mGy/h (green) irradiated HUVECs. The average ratios of relative protein expression in control and irradiated samples are shown. The protein bands were quantified using TotalLab TL100 software by integration of all the pixel values in the band area after background correction and normalized to the actin expression. The data are represented as ± SEM. Three biological replicates were used in all experiments. (Students t-test; *p<0.05, **p<0.01 and ***p<0.005).
A panel of senescence-related proteins was analyzed by RPPA; these are listed in
As the PI3K/Akt pathway is a known regulator of endothelial senescence
Expression of some proteins shown to be affected at a higher dose rate of 4.1 mGy/h in our previous study
The columns represent protein levels of phospho to total ERK ½ (A), Rho GDI (B), caldesmon (C), nucleolin (D) and SUMO1 (E) in control (blue), 1.4 mGy/h (orange) and 2.4 mGy/h (green) irradiated HUVECs. The average ratios of relative protein expression in control and irradiated samples are shown. The data are represented as ± SEM. Three biological replicates were used in all experiments. (Students t-test; *p<0.05, **p<0.01 and ***p<0.005).
Among other proteins showing differential regulation in the RPPA analysis were caldesmon, showing down-regulation of expression with the 2.4 mGy/h dose rate at weeks 6 and 10 (
Global proteome changes were assessed by ICPL triplex methodology at week 10, the time point showing the most alterations with RPPA. The ICPL triplex labeling of control and irradiated samples followed by LC-ESI-MS/MS analysis identified 2,607 proteins in total. Among 35,546 peptides the number of lysine-containing peptides was 22,587; of these 21,092 peptides were ICPL labeled which corresponds to a labeling efficiency of 93.4%. Quantification of proteins showed that 130 proteins were significantly deregulated at the 1.4 mGy/h dose rate. Of these, 50 were down-regulated and 80 up-regulated (Table S1 in
The networks where the significant deregulated proteins were involved were analyzed by the IPA software; the biological pathways that were significantly altered by the chronic radiation exposure are shown in
# | Top canonical pathway | p-value | Ratio |
1.4 mGy/h dose rate | |||
1 | EIF2 signaling | 8.51E-13 | 16/174 (0.092) |
2 | Regulation of EIF4 and p70S6K signaling | 1.67E-04 | 7/151 (0.046) |
3 | Remodeling of epithelial adherens junctions | 2.25E-04 | 5/65 (0.077) |
4 | Arginine biosynthesis IV | 7.05E-04 | 2/5 (0.4) |
5 | Glycogen biosynthesis II (from UDP-D-Glucose) | 7.05E-04 | 2/5 (0.4) |
2.4 mGy/h dose rate | p-value | Ratio | |
1 | EIF2 signaling | 6.71E-67 | 61/174 (0.351) |
2 | mTOR signaling | 6.68E-22 | 30/185 (0.162) |
3 | Regulation of EIF4 and p70S6K signaling | 1.82E-20 | 26/151 (0.172) |
4 | Remodeling of epithelial adherens junctions | 1.14E-05 | 8/65 (0.123) |
5 | Caveolar-mediated endocytosis signaling | 1.56E-04 | 7/78 (0.09) |
Shared pathways between the two dose rates | p-value | Ratio | |
1 | EIF2 signaling | 2.45E-13 | 13/174 (0.075) |
2 | Remodeling of epithelial adherens junctions | 8.64E-06 | 5/65 (0.077) |
3 | mTOR signaling | 1.01E-03 | 5/185 (0.027) |
4 | Regulation of EIF4 and p70S6K signaling | 2.89E-03 | 4/151 (0.026) |
5 | Epithelial adherens junction signaling | 3.29E-03 | 4/142 (0.028) |
Differentially regulated proteins at week 10 were uploaded to IPA. The most important canonical pathways from both dose rates are shown. Shared pathways between the two dose rates are indicated. Statistical significance (p-value) was calculated using Fischer's exact test. The ratio represents the number of identified deregulated proteins divided by the number of total proteins in that pathway.
The most affected pathway identified at both dose rates was EIF2 (elongation factor 2) signaling, a critical point in regulation of protein synthesis. The number of deregulated proteins involved in this pathway increased with the increasing dose rate (indicated in
A second pathway affected mostly at the higher dose rate, but also appearing in the shared pathway analysis (
The biological functions of the proteins affected by chronic radiation were analyzed by IPA; these, together with the corresponding proteins, are listed in the Table S5 in
The two networks representing the most important biological functions changed at the 2.4 mGy/h dose rate are shown in
Differentially up- or down-regulated proteins are marked in red and green, respectively. Nodal molecules Akt and ERK 1/2 (blue color) were predicted by the IPA software as central transcriptional regulators.
Significant deregulated proteins (2.4 mGy/h) were also categorized based on their cellular location or metabolic function (Table S6 in
There is increasing concern about low-dose and low-dose-rate radiation exposure impacts upon human health. It is possible that biological mechanisms responsible for the low-dose-rate effects differ from those of high-dose-rate effects. In radiation-induced CVD, high doses cause overt damage to the myocardium
We have previously shown that chronic radiation given at the dose rate of 4.1 mGy/h was able to induce premature senescence in primary HUVECs
We have now studied the effects of the lower dose rates of 2.4 and 1.4 mGy/h, and investigated their influence on the senescence status of HUVECs. The higher dose rate (2.4 mGy/h) was able to induce premature senescence based on the appearance of classical hallmarks: the inhibition of growth, and increased number of cells stained by SA-ß-gal. as well as increased expression of the cell cycle inhibitory cyclin-dependent kinase inhibitor p21, a known senescence marker in cellular
A potential signaling pathway involved in the initiation of senescence is indicated by the observed biphasic increase in the level of total Akt (weeks 1 and 10) accompanied by a corresponding decrease of the phosphorylated form (Akt-Ser473) at weeks 6 and 10. In parallel, the levels of PI3K (weeks 6 and 10) and mTOR (week 10) were significantly decreased.
Akt is a mediator of the stimulating effects of vascular endothelial growth factor (VEGF) on endothelial survival and migration
We compared our data to previous studies examining acceleration of senescence and proliferation arrest of endothelial precursors and mature endothelial cells in response to different stimuli
mTOR in its complex 2 form regulates actin polymerization and is thus necessary for actin cytoskeleton formation
Actin cytoskeleton and components of actin-assembly machinery, such as actin-related protein 2/3, are interacting with extracellular matrix receptors such as integrins
We found a significant down-regulation of Rho GDI with both dose rates at weeks 6 and 10. This down-regulation was, however, not as pronounced as in our previous study when a higher dose rate of 4.1 mGy/h was used
The Rho pathway, in addition to the cytoskeleton, controls the progression of the G1 phase of cell cycle to the S phase
Our study also shows the influence of chronic low-dose-rate radiation on mitochondrial proteins. We have previously shown that, in the case of acute immediate radiation damage, mitochondrial proteins represented the protein class most sensitive to ionizing radiation
This study highlights the participation of PI3K/Akt/mTOR pathway inhibition in the premature endothelial senescence triggered by chronic low-dose-rate radiation. A schematic presentation of this pathway and its downstream targets is shown in
Orange and green colors represent down-regulated proteins detected by RPPA and ICPL triplex proteomic methods, respectively. Blue color indicates unregulated proteins detected by the RPPA technique.
(XLSX)
The authors acknowledge Stefanie Winkler for technical assistance.