Background
Promyelocytic leukemia protein (PML) is part of a large multiprotein nuclear complex known as promyelocytic leukemia nuclear bodies (PML-NBs), which is associated with the nuclear matrix [
1]. Promyelocytic leukemia protein is involved in multiple cellular functions including transcription, chromatin dynamics, oncogenesis, posttranslational modifications and DNA damage response [
2,
3]. Additionally, PML plays a role in senescence and ageing. PML is upregulated during cellular senescence [
4] and induces permanent cell cycle arrest via p53 and retinoblastoma protein regulation [
5]. At the same time, PML is a dynamic sensor of DNA damage and cellular stress [
6] and it is involved in a later step of the DNA double-strand break (DSB) repair [
7]. In response to DNA damage, the number of PML-NBs increases and these PML-NBs alter their subnuclear location [
7]. One type of cellular stress is ionizing radiation due to its DNA damaging properties. The number of DSBs identified as γH2AX foci in peripheral blood monocytes is measured in order to assess the stress triggered by irradiating the cells. γH2AX foci are formed at the sites of DSB by the phosphorylation of the H2A histone variant H2AX at its amino acid serine 139 [
8]. The ratio of visible γH2AX foci to DSBs is close to 1:1, what makes the γH2AX immunofluorescent staining the most sensitive method available for detecting DSBs in human cells [
9]. Individual differences in DSB repair can be obtained from interindividual disparities in the number of γH2AX foci per nucleus after
in vitro irradiation and the subsequent repair. The highest numbers of double strand breaks per nucleus are visualized 30 minutes post irradiation. The DSB number declines over a period of hours, while the cells are repairing the DNA damage.
We evaluated the difference in patients’ PML response to stress factors like cancer and radiotherapy treatment dependent on age. We investigated the number of PML-NBs in peripheral blood mononuclear cells of healthy individuals and cancer patients of different ages. Pre-existing and via ionizing radiation induced PML-NBs were studied. As a measure of cellular stress, we compared the PML-NBs with the number of DNA double-strand breaks.
Method
Patients characteristics
The prospective study included a total of 175 individuals. 66 rectal cancer patients (RC), 68 breast cancer patients (BC) and 41 healthy participants were enrolled. The trial was in compliance with the WMA Declaration of Helsinki - Ethical Principles for Medical Research Involving Human Subjects. All patients and healthy individuals gave their written informed consent. This study was approved by the ethics review committees of the Friedrich-Alexander-Universität Erlangen-Nürnberg (No. 2725). The patients’ blood samples were taken immediately before the first irradiation fraction to yield the pre-existing background rates. After a daily fractionated radiation treatment (5 × 1.8 Gy) and a free interval of three days the second blood sample was taken (in vivo irradiation). RC patients were treated with 5-fluoruracil (5-FU) or a combination of 5-FU and oxaliplatin once a week. BC patients received docetaxel, cyclophosphamide, epirubicin or paclitaxel or a combination of these substances. Some BC patients additionally received an antihormonal therapy.
Blood samples and peripheral blood mononucleated cells separation
The blood samples taken from patients prior to radiochemotherapy (RCT) and the blood samples taken from control group were split into three samples. Two different doses had to be used to irradiate the peripheral blood mononucleated cells, because a dose of 2 Gy would induce after 30 minutes such a high amount of foci so that they could not be separated from each other and counted properly. There would remain only a very low number of foci 24 h after the low dose of 0.5 Gy and therefore the statistics would be poor. Though the first sample was in vitro irradiated using 0.5 Gy and afterwards incubated for 30 minutes. This procedure was adopted for the second sample with the difference of using 2 Gy and having a 24 h incubation time. The third sample served as control. After five fractions in vivo irradiation and the free interval of three days, the fourth sample was taken. Two identical cover slips were produced from each sample. Briefly, peripheral blood mononucleated cells (PBMC) were isolated from heparinized whole blood samples by Ficoll gradient centrifugation. PBMC were maintained in RPMI 1640 medium supplemented with 1% penicillin-streptomycin and 10% fetal calf serum. PBMC were divided into three samples, two were in vitro irradiated and the third sample was used as a control. Afterwards, the blood samples were cytocentrifuged (StatspinCytofuge, Kreatech, Germany) onto specimen. The PBMC were fixed for 30 minutes in methanol and for one minute in acetone before they were washed for 3x10 minutes in a phosphate-buffered saline with foetal calf serum.
Primary fibroblasts
A 2-mm punch skin biopsy was taken from the forearm of a healthy Caucasian individual. The dermis was cut into small pieces and placed in a small flask where it was covered with F12 medium (Biochrom, Berlin, Germany) containing 20% fetal calf serum. The outgrowing fibroblasts were trypsinized and sub-cultured.
Antibodies and immunofluorescence analysis
One of the two identical cover slips of each approach was then incubated with the mouse anti-γH2AX antibody (Abcam, Cambridge, UK) and the second cover slip with the rabbit anti-PML antibody (Santa Cruz, CA, USA). Afterwards the cover slips were washed in PBS three times and then incubated with a secondary goat anti-mouse labeled with Alexa 488 fluorescent antibody or goat anti-rabbit labeled with Alexa 594 fluorescent antibody (Molecular Probes, Karlsruhe, Germany). Then, the samples were washed in PBS again three times and mounted by using the Vectashield mounting medium (Vector Laboratories, Peterborough, UK).
Fluorescence labeled blood cells were visualized by a fluorescence-microscope (Axioplan 2, Zeiss, Göttingen, Germany) and image acquisition software (Metafer 4, MetaSystems, Altlußheim, Germany). Digital images of five optical planes separated by a distance of 0.75 μm were recorded and combined to an extended focus image using the maximum intensity algorithm (Metasystems). An area of 2 mm
2 (630×) was captured automatically. For each of the samples 500 to 1000 cells were identified by using image analysis software (Biomas, Erlangen, Germany). All nuclei were morphologically considered by eye to be properly shaped and cells in cell-division phase were excluded. By using Biomas, the PML foci and the γH2AX foci inside each nucleus were counted [
10]. The number of mean residual foci per cell was determined for every individual before irradiation, 30 minutes after a dose of 0.5 Gy and 24 h after a dose of 2 Gy. After the five fractions
in vivo irradiation and the free interval of three days the mean number of foci per cell was counted again.
In vivoexposed dose
We evaluated two quantities to consider dose aspects: The total deposit energy E
dep and the mean dose D
mean, which is the dose averaged over the whole body of the patient. These quantities are strongly connected by a proportionality relation:
where m is the mass of the patient.
The total deposit energy E
dep is estimated (here) by:
where ρ is the mass density of the patient (ρ was approximately set to 1 kg/dm3), Dpres. is the prescribed dose and Vp is the isodose volume with at least a percentage p of the prescribed dose. We used p in the following percentage steps: 0.2, 0.3 0.4, 0.6, 0.8, 0.9 and 0.95. The different volumes Vp are achieved from the dose distribution in the patients given by the treatment planning system (Pinnacle, Philips, Fitchburg, WI, USA).
Statistical analysis
The independent t-test was used to test for statistical differences. With p < 0.05, differences were considered statistically significant. Pearson correlation was used to evaluate a possible correlation between PML-NBs frequency per cell and time after exposure and between age and focus formation. Statistical calculation was performed by using SPSS version 19 (IBM, Ehningen, Germany).
Discussion
One of the major causes of cell aging seems to be impaired repair of DNA damage as a result of changes in cellular stress response which lead to an accumulation of DNA damage [
11]. However, there must be several factors leading to dysfunctional DNA repair in aged cells. We were interested in PML because of its involvement in DNA damage repair and stress response. Our major findings are: (i) PML-NBs decrease in PBMC with increasing age of individuals. At the same time the cellular stress increases, which was proved by an accumulation of γH2AX foci in dependence of age. (ii) PML-NBs arise in individuals younger than 50 years after exposure to
in vivo ionizing radiation. As opposed to this, individuals over 70 years of age show a decrease in the number of PML-NBs as a response to
in vivo ionizing radiation. There seems to be an impaired PML-NBs stress response in the aged cell. To our knowledge, so far no clinical data exist which depict the age related PML-NBs induction and a limited PML-NB stress response in older individuals.
The impaired stress response in aged cells might be related to the accumulation of DNA damage [
12]. The DNA repair is related to PML-NBs in several ways [
1]. Gamma-irradiation results in the recruitment of p53 to PML-NBs as one aspect of PML-NBs being involved in DNA repair through several repair proteins [
6]. PML-NBs are reported to facilitate the homology-directed repair by interacting with DNA repair proteins such as Rad51 or BML. It makes DSB repair more efficient by facilitating to localize and stabilize Rad51. It is suggested that PML-NBs are involved in processing double-strand breaks, generating ssDNA tails, which are essential for the assembly of DNA repair protein complexes [
2]. PML-NBs may have the ability to act as sensors of cellular stress [
6]. Although it might be that the accumulation of γH2AX foci is in part a result of the decreased and impaired stress response of the PML-NBs. Our data show that there is a relation between the aging of the cell and the decrease of PML-NBs and it might be related to the change in stress response.
We compared the pre-existing number of PML-NBs per cell in patients suffering from cancer and healthy controls. There was no significant difference in the number of PML-NBs between both groups. Therefore, the decline in the number of PML-NBs seems not to be related with the presence of cancer. Having compared the PML-NBs level in individuals at different ages, we made the observation that increased levels of γH2AX foci coincide with a decrease of PML-NBs in aged cells.
An additional support of the thesis that there seems to be an impaired PML-NBs regulation in dependence of age is that in patients younger than 50 years there was a trend to an increase of PML-NBs after a in vivo irradiation. The irradiation consisted of 1.8 Gy daily, five times for one week and three additional days. Quite in contrast to it the individuals between 50 and 70 years old had no change in the number of PML-NBs. Patients older than 70 years even had a reduced number of PML NBs after in vivo irradiation.
The PML-NB increase is caused by transcriptional upregulation [
13] or fission [
3], which results in clear and stable foci, which were made visible by immunostaining. The equally bright PML foci are easy to quantify by high throughput analyses. There is a clear difference of the PML-NBs number between lymphocytes and fibroblasts. Lymphocytes have about 2.5 PML-NBs per cell, while fibroblasts have about 10 PML-NBs per cell. In both cell types PML-NBs rapidly increase in the first three hours after exposure to ionizing radiation. PML-NBs in the lymphocytes decrease very slowly thereafter. In contrary in fibroblasts there is long lasting increase of PML-NBs after stress induction. It might be of higher priority to study the more prominent effect in fibroblasts. However, human PBMC were used for the testing. The advantage of PBMC is that all cells are in G0/G1 phase of cell cycle and so there are no problems arising from the variation of PML-NBs number depending on cell cycle. The disadvantage of PBMC is the low number of PML-NBs and the cells spheroid shape and the resulting problems to count the foci in different planes of the cell. However this can be overcome by acquiring multiple planes by the microscope and fuse it to a single image. Additionally it would have been difficult to have skin biopsies of such a large number of donors and it would not have been possible to yield skin biopsies after one week of RCT. It could be assumed that the presence and alteration of PML-NBs is similar in all cells of the body.
It could be hypothesized that PML monitors a basic aging process and not the effects of a disease [
14]. To prove this, we showed that cancer patients have no significant difference in the number of PML-NBs compared to healthy individuals. Since the accumulation of DNA damage is one of the main characteristics of aging [
15], it can be assumed that aging might be in some way involved in the DNA repair process. The roles of PML as a sensor of cellular stress and a mediator of DNA damage response is strongly implied by a series of observations made in past studies. PML-NBs were found to be sensitive detectors of cellular stress [
3]. Between PML-NBs and DNA repair links have been made, e.g. PML is supposed to have a role in coordination of DSB processing [
1]. To support this hypothesis it might be useful to study PML-NBs in animal aging models in the future [
16,
17].
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
The experimental work and analysis was carried out by BW, MS, MB, MS. BW and LD wrote the manuscript. LD and RF designed the study. All authors read and approved the final manuscript.