- Research article
- Open Access
- Open Peer Review
Necdin modulates proliferative cell survival of human cells in response to radiation-induced genotoxic stress
© Lafontaine et al.; licensee BioMed Central Ltd. 2012
- Received: 22 July 2011
- Accepted: 23 May 2012
- Published: 12 June 2012
The finite replicative lifespan of cells, termed cellular senescence, has been proposed as a protective mechanism against the proliferation of oncogenically damaged cells, that fuel cancer. This concept is further supported by the induction of premature senescence, a process which is activated when an oncogene is expressed in normal primary cells as well as following intense genotoxic stresses. Thus, deregulation of genes that control this process, like the tumor suppressor p53, may contribute to promoting cancer by allowing cells to bypass senescence. A better understanding of the genes that contribute to the establishment of senescence is therefore warranted. Necdin interacts with p53 and is also a p53 target gene, although the importance of Necdin in the p53 response is not clearly understood.
In this study, we first investigated Necdin protein expression during replicative senescence and premature senescence induced by gamma irradiation and by the overexpression of oncogenic RasV12. Gain and loss of function experiments were used to evaluate the contribution of Necdin during the senescence process.
Necdin expression declined during replicative aging of IMR90 primary human fibroblasts or following induction of premature senescence. Decrease in Necdin expression seemed to be a consequence of the establishment of senescence since the depletion of Necdin in human cells did not induce a senescence-like growth arrest nor a flat morphology or SA-β-galactosidase activity normally associated with senescence. Similarly, overexpression of Necdin did not affect the life span of IMR90 cells. However, we demonstrate that in normal human cells, Necdin expression mimicked the effect of p53 inactivation by increasing radioresistance.
This result suggests that Necdin potentially attenuate p53 signaling in response to genotoxic stress in human cells and supports similar results describing an inhibitory function of Necdin over p53-dependent growth arrest in mice.
The maintenance of genomic integrity relies on the ability of the p53 tumor suppressor to arrest the cell cycle thereby allowing correct repair of potentially oncogenic DNA damage after an insult . This checkpoint is central to prevent the accumulation of mutations in cells, which could result in carcinogenesis. Another level of protection preventing carcinogenesis is cellular senescence, a process that also involves p53 [2, 3]. Replicative senescence in human cells results from telomere shortening as a consequence of each cell division . This naturally occurring process could be related to the aging of mammals as the accumulation of senescent cells may contribute to reduce tissue functionality and may affect its morphology . On the other hand, stress-induced senescence also defines a permanent growth arrest that is triggered by irregular signaling in a cell caused by an activated oncogene or an unresolved genotoxic stress [5–8]. Replicative senescence and stress-induced senescence result from common mechanisms in which the tumor suppressors Rb and p53 play a central role . Senescent cells remain metabolically active, with a flat morphology and are characterized by β-galactosidase activity . Some effectors of p53 such as the inhibitor cyclin-dependent kinase p21, miR34 and PML are well characterized for their involvement in this permanent growth arrest [11–13]. However, new targets of p53 are continuously discovered and require extensive characterization to entirely understand their functions in the p53 pathways that regulate cell cycle, apoptosis and senescence. Moreover, p53 regulation is complex and remains incompletely understood .
Necdin has been recently identified as a p53 target gene [15, 16]. We initially observed that Necdin is increased in polyomavirus large T-antigen expressing NIH3T3 cells, a mouse model used to unravel early events in carcinogenesis . In this model, the increase in Necdin results from a p53-independent mechanism suggesting that other mechanisms are involved in Necdin transcriptional regulation. Necdin was first described as a growth suppressor with an Rb-like activity by interacting with E2F1 to repress its function [17–19]. However, we observed that NIH3T3 cells could grow efficiently even with high Necdin expression . The role of Necdin in cancer remains poorly defined. A decrease in expression is observed in melanomas , while pancreatic cancer presents increased expression through the loss of imprinting in the Necdin gene . We hypothesized that Necdin expression could be associated with better outcomes, as suggested since Necdin is associated with a better prognosis in breast cancer  and by our previous results revealing that Necdin expression is limited to borderline ovarian cancer, which is usually p53 wild type cancer .
Necdin is linked to p53 pathways suggesting that it may impact cancer development. The initial evidence by Taniura and al. , demonstrated a possible interaction between Necdin and p53. It was also reported that Necdin has the ability to suppress p53-induced apoptosis [19, 23]. We previously observed that Necdin expression allowed cells to overcome the growth arrest induced by p53 activation . In addition, Necdin can affect posttranslational modification of p53 by interacting with Sirt1 . Sirt1 is a deacetylase  and the interaction of Sirt1 with Necdin potentiates its capacity to decrease p53 acetylation leading to the reduction of p53 activity . Therefore, Necdin seems to be closely related to p53 since it is a p53 target gene and it can negatively modulate p53-dependent growth arrest and apoptosis. Its relationship to p53 activity needs to be further clarified. These observations prompted us to explore the possible impact of Necdin during senescence, another important role of p53 in preventing cancer. Here we demonstrate that while Necdin expression levels decline in both replicative and premature senescence, modulation of its expression does not affect human primary cell life span. However, high levels of Necdin contribute to an increased radioresistance in primary cells. Taken together, these observations are consistent with an inhibitory effect of Necdin on the p53 pathway and suggest a role for Necdin, under stress conditions, in preventing senescence induction.
Low passage IMR90 normal human fibroblasts  were a generous gift from Dr. Christian Beausejour (Centre de recherche du CHU Ste-Justine, Montreal). Primary cells were kept at 37°C under 5% CO2 and a low O2 condition (5%), and were maintained in Dulbecco’s modified eagle medium (DMEM) supplemented with 10% FBS. The number of living cells was determined using the CASY® cell counter model TT. Proliferation of IMR90 cells was assessed by population doubling according to the following formula: last PD + (Log (final cell number) - Log (initial cell plated) x 3.32).
Vectors and infection
All constructs for lentivirus production were derived from a previously described expression system . The human NDN gene was excised from pOBT7 containing full length human NDN (Open Biosystems, MHS1011-61084) and inserted in the 686–1 vector (pENTR4 no ccdB, Addgene number 17424). Recombination of this construct was performed with the destination expression vectors 685–3 or 670–1 (pLenti CMV/TO Neo DEST and pLenti CMV/TO Puro DEST: Addgene number 17292 and 17293) using the Gateway LR Clonase® enzyme mix (Invitrogen). Control vector was generated by recombination of the empty pENTR4 no ccdB with the same destination vectors. GSE22 (encoding an interfering p53 fragment) have been previously described . H-RasV12 in 685–3 vector was a gift from Christopher Wiley (from J. Campisi’s lab). pLKO.1 lentiviral shRNA vectors targeting human NDN gene (shNdn1 (TRCN0000020085), shNdn2 (TRCN0000020086)) or GFP as a control (shGFP (RHS4459)) were purchased from Open biosystems.
Lentiviruses were produced by co-transfection of the different pLenti contructs together with ViraPower Lentiviral Packaging Mix (Invitrogen) in the 293FT packaging cell line. 72 hrs later, supernatants were collected and viruses were concentrated by ultracentrifugation. Infections were performed on 5–7.5 x 105 cells overnight in the presence of polybrene. Appropriate selection was applied 48 hrs later.
IMR90 cells expressing the tetracycline repressor (TetR) were generated by infection with lentiviruses containing the 716–1 vector (pLenti-CMVtetR Blast, Addgene #17492).
Senescent IMR90 cells were generated by irradiating cells at 20 Gy in a Gammacell irradiator. Cell extracts were harvested at indicated times for western blot analysis. For oncogene-induced senescence, IMR90 cells were infected with oncogenic RasV12 containing lentivirus. Senescence was assessed by senescence-associated ß-galactosidase (SA-ß-gal) staining  using a Senescence Detection Kit (BioVision) in 6-well or 12-well culture plates according to the manufacturer’s instructions. Cells were plated 24 hrs before staining.
Growth arrest and FACS
For serum starvation, cells were washed 24 hrs after plating and medium was replaced by DMEM containing 0.1% FBS. After 24 hrs of exposure to low or normal serum conditions, cells were collected, fixed with ethanol and stained with propidium iodide. DNA content analysis was performed using a Coulter EPICS XL-MLC Flow Cytometer.
Cells were lysed with a buffer containing 50 mM Tris HCl, pH 7.4, 150 mM NaCl, 1 mM EDTA, 1% TRITON™ X-100 and protease inhibitors (Complete Protease Inhibitor Cocktail Tablets, Roche). Western blot analyses were done on nitrocellulose membranes hybridized with various antibodies: from Santa Cruz p53 (DO-1, sc-126), p21 (F5, sc-6246), H-Ras (F-235, sc-29), p16 (JC-8, sc-56330), PCNA (FL-261, sc-7907), from Millipore (Upstate) Necdin (07–565) and from Abcam β-actin (AC-15, ab6276). Secondary HRP-conjugated antibodies were all purchased from Santa Cruz.
One day after 1 x 105 cells were seeded in 6-well plates they were treated for 24 hrs with 2.5 μM Nutlin-3 (Sigma) or DMSO for the untreated control. Cells were collected and total protein extracts were performed in the lysis buffer as described above. p53 accumulation was assessed by Western blot.
Colony formation assay
For stress-induced senescence analysis, we first transduced IMR90 cells containing the tetracycline repressor (IMRtetR) with the vector of interest (empty vector and hNdn) and IMR90 cells with GSE22 or shRNA (shNdn1, shNdn2 and shGFP). After an appropriate period of selection, cells were exposed to irradiation. For each population, 100 untreated cells were seeded to determine plating efficiency. 1.6x103 cells irradiated with a dose of 2 Gy were plated in 100 mm. All conditions were performed in triplicate. Medium was replaced every third day and cells were stained after 12 days with a crystal violet solution. Percentage of colonies was determined by the following formula: number of colonies in irradiated cells/(number of colony in untreated cells x dilution factor). Since Necdin expression was inducible, we chose to maintain all populations in this experiment in medium containing doxycycline to avoid variation.
Necdin and p53 expression are linked in human fibroblasts
Necdin level decreases with replicative senescence
Premature senescence is accompanied by a decrease in necdin level
Another method to induce premature senescence in human normal cells is exposure to stresses causing persistent DNA damages . In order to induce stress-induced premature senescence (SIPS), cells were exposed to gamma irradiation. Cells that were irradiated at 20 Gy progressively entered a senescence state, which was marked by an increase in p16INK4A expression over time (Figure 3B). This increase in p16INK4A was accompanied by a decrease in Necdin protein levels (Figure 3B). Finally, we evaluated whether the decrease in Necdin levels observed with senescence was also seen in transient growth arrest. Cells were subjected to serum starvation for either 24 or 48 hours, which induced a cell cycle arrest in G1 as confirmed by DNA content (Figure 3C). Western blot analysis revealed a constant level of Necdin protein during this transient growth arrest (Figure 3D).
Thus, during both oncogene-induced (RasV12) and stress-induced (irradiation) senescence, Necdin levels decreased with the establishment of senescence. The decrease in Necdin expression is specific to permanent growth arrest since serum depletion did not generate the same effect.
Modulation of necdin does not affect replicative aging
Necdin overexpression provides radio-resistance in human normal cells
Low necdin expression as a marker of senescence
Replicative senescence marks the end of the proliferative state in cells when telomeres reach a critical length . This mechanism is relevant in limiting aberrant proliferation and may contribute to cancer prevention . By comparing IMR90 cells at high or low population doublings, we demonstrate a reduction of the Necdin protein levels over time, while p16 increases, which is the mark of an irreversible arrest [29, 37, 38]. Lee and al.  identified Necdin among the gene expression profiles in skeletal muscle of aging mice, where Necdin levels decrease with age. In contrast, Necdin was upregulated upon caloric restriction, a process that retards aging in mice . Our result confirmed that the decreased Necdin levels seen in aged mouse skeletal muscle can be reproduced with normal human fibroblasts in culture. The decrease in Necdin expression was also consistently observed in telomere-independent premature senescence induced by irradiation. It was also observed, but to a much lower extent, in premature senescence resulting from the expression of the activated oncogenic Ras (RasV12). Although the observed reduction in Necdin protein was subtle, another group has independently reported that Necdin mRNA was part of the list of downregulated genes following Ras-induced senescence , supporting our general finding that Necdin is downregulated during cellular senescence.
In contrast to senescence, cellular quiescence is a reversible cell cycle arrest since cells may re-enter the cell cycle when the restriction is removed [29, 37]. Interestingly, in a transient G1 arrest, Necdin remained expressed, suggesting that the decrease in Necdin is limited to permanent growth arrest and may be a marker of senescent cells. Necdin levels in other primary cells need to be analyzed since the molecular signature at senescence largely depends on the cell type .
Like Necdin, many genes involved in cell cycle progression show altered expression in senescent cells as both proliferation-promoting proteins and their negative regulators decrease when human fibroblasts reach senescence, as exemplified by Rb and E2F1 [34, 41]. Necdin interacts with E2F1, E2F4 and also affects expression of the Rb family of genes . One interpretation of these findings is that Necdin may affect cellular senescence. To test this hypothesis, ectopic expression of Necdin was used. We found that overexpression of Necdin did not extend the replicative life span of primary human fibroblasts. Moreover, the sustained expression of Necdin over time in this experiment suggested that the complete elimination of Necdin expression is not essential for senescence to occur.
There are various examples in the literature where proteins, possessing properties similar to Necdin, can induce premature senescence when depleted in IMR90 cells. The knock down of BS69 and p400, also known to interact with viral proteins, induced a premature senescence characterized by p21 and p53 upregulation [43, 44]. BS69 and p400, like Necdin, can form a complex with p53 at the p21 promoter to repress transcription [19, 44], although it is not known if Necdin is part of the same complex. Moreover Sirt1, a partner of Necdin , can increase the risk of cancer due to its capacity to downregulate p53 activity by deacetylating this tumor suppressor [24, 45, 46]. Accordingly, Sirt1 inhibition also induces a senescence-like phenotype in IMR90 cells . From these data, we expected that knock down of endogenous Necdin by shRNA might also induce premature senescence resulting from p53 activation and p21 de-repression. However, shNDN-directed loss of Necdin expression did not induce premature senescence in IMR90 cells. Perhaps, under non-stress conditions, a complementary protein may contribute to maintaining Necdin function in its absence. For example, other members of the MAGE family like NDNL2, also known as MAGE-G1, , are expressed in a wide variety of tissues including fibroblasts  and shares many functions with Necdin. Alternatively, loss of Necdin might not be sufficient to activate p53 in the absence of others p53 stabilization signals.
Necdin is a maternally imprinted gene and its promoter contains many CpG sites for methylation regulation . Consistent with this, the inhibition of DNA methyltransferase (DNMT) by 5-aza-2′-deoxycytidine (5AZA-dC) has been shown to induce Necdin expression in some cancerous cell lines . During senescence establishment, epigenetic changes occur inducing important chromatin structure modifications; some at specific sites while other reflect a more global change. Global DNA methylation status decreases with aging by comparison to young counterparts and immortalized cells [52, 53]. A specialized redistribution in chromatin heterochromatin, called senescence-associated heterochromatic foci (SAHF), is also associated with cellular senescence . The mechanism is only partially understood; the genomic loci affected by this structure often contain proliferation-promoting genes such as E2F1 target genes . Moreover, SAHF are not observed in quiescent cells. It is possible that the decrease in Necdin levels during aging could be the result of hypermethylation or others senescence associated modifications at specific sites.
Necdin expression confers resistance to ionizing irradiation
Necdin expression can exert an effect on normal cells as an increase in Necdin level confers resistance to ionizing radiation. At the employed dose, DNA double-strand breaks were induced in all cells and the physiological consequence is the induction of a DNA damage response activating p53. Thus, these damages caused cell cycle arrest and need to be repaired before cells can resume proliferation. When damages cannot be properly repaired, apoptosis or senescence are two possible cellular outcomes. Consequently, we observed that cells unable to form colonies showed enlarged and flat morphology of senescent cells in all populations tested (data not shown). This is expected for normal fibroblasts that are relatively resistant to apoptosis upon irradiation . Apoptosis was probably also induced to a low level but could not be monitored due to the low cell densities used in these experiments.
An appropriate response to genotoxic stress is based on the capacity to sense the damage, to activate the cell cycle checkpoint and to repair the damage . Two principal conditions may explain an increase in resistance to irradiation in a normal cell. First, an enhanced ability to repair the DNA damage could promote survival. Second, a failure in activating cell cycle checkpoints will contribute to the maintenance of cells in a proliferative state despite the presence of a genotoxic stress. This could result from inefficient sensing upstream of the p53 pathways or by a reduction in p53 downstream signaling. This is what was reproduced with the positive control expressing a peptide inhibitor of p53 (GSE22) resulting in a marked increase in colony formation consistent with the notion that a decrease in p53 function increases resistance to irradiation. We observed and others have shown previously that Necdin can interfere with p53-responses [16, 19, 23]. These data suggest that increased radioresistance associated with Necdin may be related to its ability to directly influence the p53-response. Necdin may also confer radioresistance by inhibiting radiation-induced apoptosis since Necdin can negatively modulate caspase activation upon genotoxic stress [56, 57], but this is unlikely in fibroblast in response to irradiation . Further evidence supporting Necdin’s ability to contribute to radioresistance is a microarray analysis of the gene expression profiles produced by radiosensitive and radioresistant esophageal carcinoma cell lines. In this study, Necdin expression was higher in radioresistant cells  which is consistent with the observation of the present study.
The results of the present study suggest that Necdin function in regulating p53 responses is revealed only under stress conditions. In the absence of exogenous genotoxic stress, Necdin has no effect on normal cellular life span in human cells.
We thank Drs. Luke Masson and Philippe Gannon for helpful discussion and comments on the manuscript. We are also grateful to Dr. Eric Campeau who generously provides us all vectors for cloning and lentivirus production. We thank Drs. C. Wiley and J. Campisi for the RasV12 vector. This research was supported by a grant from the Canadian Institutes of Health Research to A.-M. M.-M. (MOP-36056). A.-M. M.-M. and F.R. are Scientists of the Centre de recherche du CHUM which receives funding from the Fonds de recherche du Québec – Santé. J.L. was supported by Canderel fund of the Institut du cancer de Montréal.