Cellular transition to senescence is associated with extensive chromatin reorganization and changes in gene expression. Recent studies appear to imply an association of lamin B1 (LB1) reduction with chromatin rearrangement in human fibroblasts promoted to senescence, while the mechanisms and structural features of these relationships have not yet been clarified. In this work, we examined the functions of LB1 and the lamin B receptor (LBR) in human cancer cells. We found that both LB1 and LBR tend to deplete during cancer cell transfer to senescence by γ-irradiation. A functional study employing silencing of LBR by small hairpin ribonucleic acid (shRNA) constructs revealed reduced LB1 levels suggesting that the regulation of both proteins is interrelated. The reduced expression of LBR resulted in the relocation of centromeric heterochromatin (CSH) from the inner nuclear membrane (INM) to the nucleoplasm and is associated with its unfolding. This indicates that LBR tethers heterochromatin to INM in cycling cancer cells and that LB1 is an integral part of this tethering. Down-regulation of LBR and LB1 at the onset of senescence are thus necessary for the release of heterochromatin binding to lamina, resulting in changes in chromatin architecture and gene expression. However, the senescence phenotype was not manifested in cell lines with reduced LBR and LB1 expression suggesting that other factors, such as deoxyribonucleic acid (DNA) damage, are needed to trigger senescence. We conclude that the primary response of cells to various stresses leading to senescence consists of the down-regulation of LBR and LB1 to attain reversal of the chromatin architecture.
One of the cellular responses to long-lasting DNA damage is transition to senescence, which is defined as a state of permanent cell cycle arrest . Cellular senescence is a cellular response to a variety of stresses [2–4]. All stresses that induce senescence [5–8] also trigger the DNA damage response-activating tumor suppressor p53 via p21 or Rb through p16INK4a . Up-regulation of p16 is particularly prevalent in benign lesions and is often lost upon malignancy . Most known senescence-associated markers were obtained in studies of oncogene-induced senescence or replicative senescence in human diploid fibroblasts [11,12]. They include enlarged cellular morphology and large senescence-associated heterochromatin foci (SAHF) enriched in heterochromatin markers, such as H3K9me3 and HP1 proteins. However, SAHF formation does not occur in all senescent cells . Previously, it was reported that lamin B1 (LB1) expression is reduced in replicative and oncogene-induced senescence, which delays cell proliferation and promotes cellular senescence via a p53- and Rb-dependent mechanism [14–17]. In addition, increased lysosome-mediated autophagy was observed in senescent cells, participating in the degradation of chromatin fragments extruded from the nucleus to the cytoplasm , the degradation of proteins of the inner nuclear membrane (INM) and lamins [19–22]. Chandra et al.  observed that areas that have lost LB1 in senescent cells are enriched by lamina-associated domains (LADs), suggesting that the loss of LB1 might be involved in the architectural changes to chromatin that they observed earlier . Independently, Sadaie et al.  showed that LB1 is preferentially depleted during senescence from the chromatin regions containing LADs enriched for H3K9me3, a characteristic for constitutive heterochromatin, which promote the formation of SAHF. In addition, these authors observed, despite the global reduction in LB1 level, the increased binding of this protein with gene-rich regions where H3K27me3, characteristic for facultative heterochromatin, increased. However, there are no known specific reader sequences of LB1 for the recognition of lysine-methylated residues of chromatin that allow attachment to the lamina. We assume that this function is executed by the receptor of lamin B1 (LBR), which is the subject of the present study.
The structure and function of LBR in the attachment of heterochromatin to the INM have been elucidated by several authors [25–29]. These authors showed that LBR is an integral protein of the INM to which it binds via its hydrophobic transmembrane domains; and via its nucleoplasmic region, it associates with heterochromatin and attaches LB1. Hirano et al.  demonstrated that LBR is a unique protein of the INM that plays a role in both heterochromatin organization and transcription repression under the INM. The Tudor domain of the LBR nucleoplasmic region selectively recognizes histone modifications, and this entire region is required for binding to heterochromatin and transcriptional repression. Globular domain II binds to HP1, and other transcription repressors, MeCP2 and LB1. Mutations in LBR cause the Pelger–Huet anomaly in humans and changes in the undifferentiated granulocytes of ichthyotic mice [30,31]. These mutations lead to the absence of functional LBR and consequent changes in heterochromatin organization. Pericentric heterochromatin in LBR-lacking cells clusters into large spots and moves away from the nuclear membrane toward the nuclear center [32,33].
LBR is an important protein that tethers heterochromatin to the INM in undifferentiated cells [28,29,34,35]; therefore, we studied its role in the attachment of heterochromatin to the INM in mammary carcinoma MCF7 and osteosarcoma U2OS cells, and in these cells promoted to senescence after γ-irradiation. We observed that as early as 24 h after irradiation, both cell lines had lost not only LBR but also LB1 expression. In parallel, the level of p21 protein increased, indicating that both these changes realized at the onset of senescence induced by γ-irradiation are important for the activation of this process. The presented results also show that characteristic markers observed in oncogene-induced senesce in normal fibroblasts [11,12] are also present in cancer cells promoted to senescence by γ-irradiation except for SAHF but strong chromatin condensation.
Material and methods
Cell culture and small hairpin ribonucleic acid plasmid transfection
Human MCF7 mammary carcinoma and U2OS osteosarcoma cells were grown in Dulbecco's modified Eagle's medium with 10% fetal bovine serum (Gibco, Thermo Fisher Scientific), 100 U/ml penicillin, and 0.1 mg/ml streptomycin (Sigma-Aldrich). All cells were grown at 37°C and 5% carbon dioxide (CO2). Plasmid-based small hairpin RNAs (shRNAs) were used for stable LBR silencing in both cell lines. Each MISSION shRNA clone (Sigma-Aldrich) was constructed within the lentivirus plasmid vector pLKO.1-puro. Five different LBR-specific shRNAs for sequence-specific LBR silencing and the empty vector TRC1.5-pLKO.1-puro, which was used as a negative control, were obtained from Sigma-Aldrich.
Transfection was performed in six-well plates using Lipofectamine LTX with PLUS reagent (Cat no: 15338030, Invitrogen). The cells were seeded the day before to reach 50–70% confluence the day of transfection. The cells in each well were transfected with one of five different shRNA clones, and the cells in one well were transfected with the negative control. A 240 µl aliquot of Opti-MEM medium (Gibco/Life Technologies) and 1 µg of plasmid DNA in 10 µl of TE buffer were added to 1.5 ml test tubes and mixed well. Then, 1 µl of the PLUS reagent from the kit (Lipofectamine LTX with PLUS, Invitrogen) was added, mixed well, and incubated for 5 min. Next, 2 µl of Lipofectamine LTX was added, mixed gently, and incubated for 30 min at room temperature (RT). The mixture was then added to cells growing in 1.75 ml of fresh standard medium in one well of the plate. After 24 h, the medium was replaced with fresh complete growth medium containing 1 µg/ml puromycin (Sigma-Aldrich) to select positively transfected cells. Every other day, the medium was replaced with freshly prepared selective medium until the fifth day after transfection. The cells were trypsinized and plated on microscope slides to determine LBR silencing efficiency by immunofluorescent detection of LBR using two different antibodies. The clones with the highest transfection efficiency and the lowest level of LBR protein were selected to generate stable MCF7-LBR(−) and U2OS-LBR(−) cell lines. The target sequence of the shRNA giving the highest inhibition of LBR expression in the majority of cells was:
Irradiation was performed using a 60Co γ-ray source at Chizostat (Chirana, CR). Cells were seeded at a density of 2 × 105/ml and irradiated 24 h later in culture medium at 37°C under normal atmospheric conditions, with either a single fraction of 2 Gy or 10 Gy (D = 1 Gy/min) or five repeated fractions of 2 Gy delivered in 24 h intervals. Cells were irradiated in 25 mm2 culture vessels to assess the level of certain proteins and mRNA expression, and on slides in four-well dishes (Nunc, #167063, Thermo Scientific, Rochester, NY, USA) for the immunodetection of DNA double strand breaks (DSBs), heterochromatin markers, lamins, LBR, and SA-β-gal (senescence-associated β-galactosidase) activity. After irradiation, the cells were incubated at 37°C and 5% CO2 until further treatment, with replacement of the growth medium every other day.
Antibodies and immunofluorescence
Cells cultured on microscope slides were withdrawn at different time intervals after irradiation, washed twice in phosphate-buffered saline (PBS; 140 mM NaCl, 2.7 mM KCl, 1.5 mM KH2PO4, and 6.5 mM Na2HPO4; pH 7.2), and fixed in 4% paraformaldehyde in PBS for 10 min at 22°C. The cells were rinsed quickly in PBS, washed three times for 5 min in PBS, permeabilized in 0.2% Triton X-100/PBS for 15 min at RT, and washed twice for 5 min. Before incubation with primary antibodies (overnight at 4°C), the cells were blocked with 5% inactivated FCS + 2% bovine serum albumin/PBS for 30 min at RT. Antibodies from two different hosts were used on each slide to detect two different antigens in the same nuclei. Anti-H2AX phosphorylated at serine 139 #05-636, anti-H3K9me3 #05-1242, anti-HP1γ #MAB3450, anti-p21 #05-345, and anti-p16 #MAB4133 antibodies were obtained from Millipore; anti-53BP1 #4937, anti-p53 #2524T, anti-phospho-p53-ser15 #9286, and anti-β-actin #4970 antibodies were obtained from Cell Signaling; anti-lamin B1 #ab8982, anti-lamin B receptor #ab32535, and anti-Rb #ab181616 antibodies were obtained from Abcam. Anti-lamin B receptor was also obtained from LSBio #LS-C105234/56705. Anti-lamin A/C #SAB4200236 was from Sigma-Aldrich and anti-p15 #sc-67279 antibody was from Santa Cruz. The secondary antibodies, affinity purified-FITC conjugated donkey anti-mouse and affinity purified Cy3-conjugated donkey anti-rabbit, were obtained from Jackson ImmunoResearch Laboratories (West Grove, PA, USA). After the slides were preincubated with 5% donkey serum/PBS for 30 min at RT, a mixture of both antibodies was applied to each slide and incubated for 1 h in the dark at RT. This incubation was followed by washing three times for 5 min each in PBS. Cells were counterstained with 1 µM TOPRO-3 (Molecular Probes, Eugene, OR, USA) in 2 × saline sodium citrate (SSC) prepared fresh from a stock solution. After the cells were briefly washed in 2 × SSC, Vectashield medium (Vector Laboratories, Burlingame, CA, USA) was used for final mounting of the samples.
Confocal fluorescence microscopy of spatially fixed cells
The immunofluorescence of the detected proteins was analyzed using images obtained with a high-resolution Leica DM RXA confocal microscope (Leica, Wetzlar, Germany) equipped with an oil immersion Plan Fluotar objective (100 × /NA1.3) and a CSU 10a Nipkow disk (Yokogawa, Japan) for confocal imaging. A CoolSnap HQ CCD-camera (Photometrix, Tuscon, AZ, USA) and an Ar/Kr laser (Innova 70C Spectrum, Coherent, Santa Clara, CA, USA) were used for image acquisition. Automated exposure, image quality control, image analysis, and other procedures were performed using Acquiarium software . The exposure time and dynamic range of the camera in the red, green, and blue channels were adjusted to the same values for all slides to obtain quantitatively comparable images. Forty serial optical sections were captured at 0.2 µm intervals (along the z-axis). In total, 100–300 cells were recorded for each set of conditions, and the experiments were repeated two or three times. The results are reported as SEM. A t-test was used for the statistical comparison of specified points.
Senescence-associated β-galactosidase assay
Detection of senescence-associated β-galactosidase (SA-β-gal) activity was performed on day 7 after cell exposure to DNA damage using a Senescence Detection Kit #K320-250 from Bio-Vision Incorporated (Milpitas, CA, USA) according to the manufacturer's instructions. Before using the kit, the cells were fixed in 4% paraformaldehyde for 10 min at RT and washed three times in PBS. Images were captured using an Olympus BX51 microscope equipped with an Olympus DP72 camera and Quick Photo Micro 2.3 software at 200× magnification. In total, 100–200 cells in each sample were blind counted for SA-β-gal positivity in MCF7 cells in two independent experiments; for this purpose, the recorded images were magnified using Adobe Photoshop.
SDS-PAGE and Western blotting
Cells were washed in PBS, scraped in the presence of Complete Mini EDTA-free protease inhibitors (Roche Diagnostics, #04693159001) and a cocktail of phosphatase inhibitors PhosSTOP (Roche Diagnostics, #04906845001), and centrifuged. After washing in PBS with a mixture of protease and phosphatase inhibitors, the cells were transferred to Laemmli SDS lysis buffer (62.5 mM Tris, pH 6.8; 100 mM DTT; 2% SDS; and 10% glycerol) supplemented with protease and phosphatase inhibitor cocktails, sonicated briefly, and centrifuged at 14 000 × g for 10 min. The protein concentration was estimated using the Bradford assay (Bio-Rad Laboratories Inc.). Bromophenol blue (0.01%) and 100 mM DTT were added to the lysates before separation in polyacrylamide gels. Twenty micrograms of total protein was loaded per lane of an 8% SDS-PAGE gel to separate lamin B1, lamin B receptor, and lamin A/C, and a 15% SDS-PAGE gel to separate other proteins. To detect these proteins in LBR(−) and LB1(−) clones, 50 µg of total protein was loaded. After electrophoresis, the proteins were transferred to polyvinylidene fluoride membranes (Bio-Rad) using semidry transfer. Proteins were stained with the specified monoclonal antibodies overnight. Detection was performed using SuperSignal West Pico Chemiluminescent Substrate Kits (#3482 Mouse IgG and #34083 Rabbit IgG detection kits, Thermo Scientific). β-actin was used as a marker of equal protein loading. The protein signals were captured using a Fujifilm LAS-3000 Imager.
Quantitative real-time reverse-transcription polymerase chain reaction
Total RNA was isolated from control and γ-irradiated MCF7 and U2OS cells using a High Pure RNA Isolation Kit #11 828 665 001 (Roche). Cells were irradiated with a dose of 10 Gy, and RNA was isolated immediately before irradiation and at 24 h, 72 h, and 7 days post-irradiation (PI). In total, 106 cells were collected for each cell sample in three independent biological replicates. cDNA synthesis was performed using Transcriptor High Fidelity cDNA Synthesis Kit #05 081 955 001 with oligo (dT)18 as a primer (Roche). PCR was performed in a 15 µl reaction mixture containing 4 µl of template cDNA, 0.75 µl of each 10 µM primer, 7.5 µl of 2× Fast Start SYBR Green Master mix (Roche, Germany), and 2 µl of distilled water. Samples from the same experiment were analyzed in the same PCR cycle in a 96-well plate format on an Applied Biosystems 7300 thermocycler. Primers for LA/C and LBR were synthesized as previously detailed (Freund et al. ). Primers for LB1 were newly designed based on the LB1 sequence (GenBank no. NM_005573): B1_for: 5′-CCCCAACTGACCTCATCT-3′ and B1_rev: 5′-TCCTCCTCCTCCTCTTCT-3′. The primer pairs anneal to a conserved exon region at positions 1526-1543 (B1_for) and 1656-1673 (B1_rev). The primers for the human GAPDH reference gene were as follows: GAPDHQF: 5′-TGCACCACCAACTGCTTAGC-3′ and GAPDHQR: 5′-GGCATGGACTGTGGTCATGAG-3′. The amplification conditions were: initial denaturation at 95°C for 3 min, followed by 40 cycles of 15 s at 95°C, 20 s at 57°C, and 30 s at 72°C. A denaturation step was introduced to verify PCR product quality. SYBR Green fluorescence was monitored consecutively after the extension step. The cycle threshold (Ct) values ranged from 16 to 25. Expression levels were calculated using MS Excel. Each sample included three technical replicates. The transcript levels of a sampled gene were normalized to the GAPDH reference gene using the equation: sample units = 2−[ct(sample) – ct(reference)]. The sample and the reference gene were amplified in parallel (same 96-well plate) which allowed direct comparison of expression levels.
We used a primary antibody directed against LA/C to detect the INM and a DNA probe complementary to the centromere-specific satellite heterochromatin (CSH) of specific chromosomes to detect these sequences. We selected LA/C to detect lamina/INM because this protein forms a continuous layer under the INM in all cells studied in this work. In contrast with LA/C, LBR is only present in wild-type MCF7 and U2OS cells, where it forms microdomains (a discontinuous layer)  overlapping with the LA/C (Figure 5B). Centromere enumeration probes, directly labeled in orange, were used for selected chromosomes (HSA): HSA4 and HSA18 are characterized by very low transcription activity and very compact chromatin , and HSA11 and HSA17 contain highly transcribed genes and decondensed chromatin; the probes were purchased from MetaSystem GmbH (Altlussheim, Germany) and consist of chromosome-specific highly repeated human satellite DNA sequences located in the centromere (CSH). The immunoFISH protocol was adopted from Pombo et al. , with slight modifications by Falk et al. .
Ionizing irradiation induces the transition of cells to senescence accompanied by changes in chromatin structure
We analyzed the changes in chromatin structure of γ-irradiated cells. First, an increased number of growing H3K9me3 and HP1γ foci showed enhanced heterochromatinization occurring from 24 h to the seventh day PI (Figure 1A–D). At later time intervals (from the seventh day after irradiation), flat nuclei, which are characteristic of senescent cells were observed (Figure 1F,G). In these cells, chromatin was largely condensed into irregular ‘cords’, densely stained with Topro 3, resembling that of normal cells just entering prophase. In contrast with SAHF, condensed chromatin in the γ-irradiated nuclei of MCF7 and U2OS cells was distributed irregularly with many regions lacking Topro 3 staining (Figure 1F,G). Second, irradiation induced the formation of numerous nuclear blebs and micronuclei (MN) (Figure 1B–F,H), and frequent chromatin bridges between daughter cells (not shown). Up to 72 h PI, the majority of blebs and MN in MCF7 cells contained low-density chromatin, LB1, LAC, and generally lacked LBR, heterochromatin markers such as H3K9me3 and HP1γ, and the CSH (Figure 1B,C,H,J). These structures probably represent euchromatin released from the nucleus, through lamina attenuated by a decreased level of LB1. They also appear in senescent cells that express SA-β-gal, a marker of senescence, at 5–12 days PI, together with another type of MN that is phenotypically distinct from the first euchromatic MN. It contains mostly heterochromatin and frequently large amounts of LBR and γH2AX in the whole volume of the MN but not 53BP1, LA/C, and LB1 (Figures 1D–F,J and 2B,D). These MN are not coated by lamins and might represent fragmented heterochromatin, sometimes attached to LBR and eliminated from the nucleus through ruptures in the LA/C. Quantification of control and senescent MCF7 and U2OS cells containing different types of MN are shown in Figures 1J and 2D. Another feature of senescent cells was a frequent disintegration of nuclei (Figures 1I and 2A) and enlarged cytoplasm with multiple long protrusions filled with emerin (Figure 2C). Emerin belongs to the LEM domain proteins of the INM and binds to the LA/C. The presence of emerin in the cytoplasm of senescent cells indicates degradation of the LA/C, which was observed especially in U2OS cells (see later).
Development of changes in chromatin structure in MCF7 cells after γ-irradiation with the dose of 5 × 2 Gy and transition to senescence
Altered nuclear morphology of U2OS cells after γ-irradiation and induction of senescence
Decreased LBR and LB1 precedes the transition of cells to senescence
We determined the influence of irradiation dose, delivery method (single or cumulative, delivered in fractionated doses), and time PI on the cellular levels of LBR and LB1 proteins. It was evident that the number of cells lacking an LBR signal increased with the IR dose and the extent of DNA damage (Figure 3A,B). At the lowest IR dose studied (2 Gy), cells lacking LBR were very rare, with no signs of senescence. After a dose of 2 × 2 Gy, the fraction of cells lacking LBRs reached 56% at 24 h, but had decreased to 29% 6 days later, when approximately 12% of cells also expressed senescence-associated SA-β-gal (Figure 3B,C). The decrease in cells lacking LBR at later times PI with this radiation dose seemed to reflect the proliferation of cells that had successfully repaired the DSBs, did not lose LBR, and could divide (Figure 3C, enlarged image).
Simultaneous decreases in LBR and LB1 induced by γ-irradiation or by LBR-specific shRNA in MCF7 and U2OS cells.
As expected, higher radiation doses (beginning with 3 × 2 Gy), resulted in increased proportions of LBR-lacking and SA-β-gal-expressing cells (Figure 3B). Irradiation with a maximum cumulative dose of 10 Gy split into five doses (5 × 2 Gy) resulted in a loss of LBR signal in 92% of cells at 24 h after the last dose, and this phenotype remained unchanged until the seventh day PI when almost all these cells started to express the SA-β-gal senescence marker (Figure 3B). By contrast, only 76% of cells showed LBR depletion at 24 h PI with a single dose of 10 Gy; compared with fractionated dose delivery, this fraction reached 100% on the seventh day PI. Cell fractions with active SA-β-gal were low at 24 h PI reaching only 4–10% after doses of 2 × 2 Gy, 3 × 2, and 4 × 2 Gy. A small increase in cell number expressing SA-β-gal 24 h after the radiation dose of 5 × 2 Gy can be accounted for by a small population of senescent cells induced by the first radiation doses because the irradiation with the dose of 5 × 2 Gy lasted for 5 days (the doses of 2 Gy were delivered in 24 intervals). These results reveal that a fraction of the LBR-negative cells was committed to senescence at 24 h PI; however, activity of SA-β-gal was only detected on day 7 PI. The loss of LBR was accompanied by a decrease in LB1, as observed by immunofluorescence at 24 h PI in both MCF7 and U2OS cells irradiated with 10 Gy (Figure 3A,D,E). Interestingly, 10–20% of U2OS cells still preserved LBR from days 7–13 PI (Figure 4A); however, the majority of this LBR was expelled from the INM and dispersed in the cytoplasm or condensed in MN with a high content of HC, which were oversaturated with this protein (Figures 2B, 3D,F, and 4B). This distribution could be related to the relatively higher (about twofold) expression of LBR in U2OS compared with MCF7 cells (discussed later in Figure 5A) and the possible accumulation of these transcripts in some cells. Cells with atypical LBR distributions (Figure 4A,B) might influence the percentage of cells with activated SA-β-gal; the fraction of cells positive for SA-β-gal was lower in U2OS (71%) than in MCF7 (90%) at day 7 PI (Figure 4C). Some U2OS cells that had already lost LBR still contained a high level of LB1 protein, which was distributed in large clumps (Figures 3D and 4B). Senescent cells also showed increased levels of phospho-p53-s15 and p21, but not p16 (Figure 3E).
Irregularities in LBR elimination from the nuclei of U2OS cells transferred to senescence after γ-irradiation, growth curves of stable clones MCF-(LBR(−) and U2OS-LBR(−) in which LBR was reduced by shRNA and expression of SA-β-gal in parental and LBR(−) cells after irradiation with 10 Gy of γ-rays
We next determined the mRNA levels of LBR and LB1 in these cells by quantitative real-time reverse-transcription polymerase chain reaction (qRT-PCR). Small but significant decreases in the LBR and LB1 transcripts were observed in both MCF7 and U2OS cells as early as 24 h after exposure to 10 Gy of γ-rays (Figure 5A). At this time point, LBR and LB1 mRNA levels decreased by 25 and 76%, respectively, in MCF7 cells. In U2OS cells, LBR and LB1 mRNA levels decreased by 18 and 14%, respectively, compared with non-irradiated cells. At 72 h PI, LBR and LB1 mRNA expression levels decreased by 62 and 81%, respectively, in MCF7 cells; and by 72 and 73%, respectively, in U2OS cells. At day 7 PI, LBR and LB1 mRNA expression levels further decreased by 69 and 87%, respectively, in MCF7 cells, and by 76 and 89%, respectively, in U2OS cells. LA/C expression was less consistent than that of LBR and LB1. In U2OS cancer cells, the level of LA/C expression did not change within 24 h PI, but decreased by 37 and 54%, respectively, at 72 h and 7 days PI (Figure 5A). By contrast, MCF7 cells exhibited a twofold increase in LA/C transcripts within the 72 h interval PI, followed by a decrease at day 7 PI. The levels of LA/C relative transcription also differed between the control cell lines (Figure 5A).
Relative expression of LB1, LBR, and LA/C after irradiation of MCF7, U2OS cells, and stable clones in which LBR was reduced by shRNA
The changes in LBR and LB1 at the transcript level were reflected by changes in the protein products (Figure 3E). On Western blots, the antibody reactivity against both LBR and LB1 proteins was always weaker in irradiated cells than in controls, which was in accordance with previous immunostaining of cells (Figure 3A,D). A low-molecular weight band corresponding to lamin C markedly increased PI (particularly at day 7) in U2OS cells, but not in MCF7 cells (Figure 3E). The increase in the lamin C band was accompanied by a decrease in the lamin A band. In addition, smeared bands migrating beneath the lamin A band were observed, especially in U2OS cells at day 7 PI, indicating protein degradation (Figure 3E).
Effect of LBR down-regulation on the induction of senescence and chromatin structure changes
Using shRNA complementary to LBR mRNA, we prepared stable clones of MCF7 and U2OS cells with reduced expression of the LBR protein [clones denoted as LBR(−)]. We analyzed the levels of LBR, LB1, and LA/C transcripts in these clones and the parental lines using qRT-PCR (Figure 5A). The LBR and LB1 transcripts were reduced by 78 and 67% in U2OS-LBR(−) and by 69 and 27% in MCF7-LBR(−), respectively. Similar reductions in protein levels were observed by immunodetection in nuclei (Figure 3A,D) and Western blotting (WB) (Figure 3E).
The phenotypic differences between the parental and LBR(−) clones were relatively minor (Figure 3A,D). Cells with reduced LBR levels exhibited slower proliferation than the parental cells (Figure 4D). Compared with parental cells, the MCF7-LBR(−) cells had a somewhat higher proportion of SA-β-gal-positive cells (Figure 4C). However, it should be mentioned that the SA-β-gal positivity in MCF7 parental cells was always higher compared with U2OS parental cells. The difference between the proportion of senescent cells in the parental control and LBR(−) cells was less pronounced in U2OS cells (Figure 4C). A significant difference was observed at day 7 PI; at this time point, SA-β-gal positivity reached 89% in LBR(−) clones and only 71% in parental U2OS cells. Similarly, the number of SA-β-gal positive cells was slightly higher (97%) in MCF7-LBR(−) clones compared with parental cells (90%) indicating the faster transition to senescence of LBR(−) clones after irradiation. SA-β-gal positivity of parental U2OS cells reached almost 100% by day 12 PI, showing their slower transition to senescence. In contrast with the parental cells, LBR(−) clones also produced more euchromatin-rich MN in senescence (Figures 1J, 2D, and 3F). Only a few U2OS-LBR(−) cells exhibited MN saturated with LBR (Figure 3F) in contrast with the parental cell line at day 7 PI (Figure 2D), similar to MCF7-LBR(−) cells (Figures 1J and 3F); the majority of cells exhibited morphological changes (large cells, giant and flat nuclei) characteristic of senescence.
At the expression level, the already low LBR levels did not change significantly after irradiation and during the transition to senescence (Figures 3E and 5A). The LB1 transcripts gradually decreased, by approximately 86 and 93%, in irradiated MCF7 and U2O2S cells at day 7, respectively. The LA/C transcripts decreased by 50% at 24 h PI and by 80 and 78% at 72 h and 7 days PI, respectively, in irradiated compared with non-irradiated U2OS-LBR(−) cells. A similar trend was observed in the MCF7-LBR(−) clone. On Western blots, fast migrating bands appeared after immunostaining with the lamin A antibody on day 7 PI, indicating the marked degradation of this protein in both MCF7-LBR(−) and especially, U2OS-LBR(−) clones (Figure 3E).
LBR attaches centromeric heterochromatin to the INM in cancer cells
We then examined the interactions between the INM (marked by LA/C, Figure 5B) and the CSH of gene-poor (heterochromatin-rich) and gene-rich (heterochromatin-poor) chromosomes using immunoFISH. For these analyses, two gene-poor chromosomes, HSA18 and HSA4, and two gene-rich chromosomes, HSA17 and HSA11, were selected. Gene-poor chromosomes are located closer to the nuclear periphery [40,41], whereas gene-rich chromosomes are closer to the nuclear center. Among all human chromosomes, HSA18 and HSA4 are the closest to the INM. Hence, we hypothesized that the location of chromosome territories could influence the frequency with which the CSH of a given chromosome interacts with the INM.
The MCF7 cells exhibited HSA4 trisomy and HSA18 trisomy or tetrasomy. U2OS cells were more heterogeneous, having four to five HSA4 and four to seven HSA18 chromosomes. We also observed HSA17 trisomy and HSA11 tetrasomy in MCF7 cells, whereas HSA17 tetrasomy and HSA11 trisomy were observed in U2OS cells. In senescent cells, the number of these chromosomes further increased by 2–4-fold, probably due to endoreduplication and polyploidy developed after irradiation of the cells (Figure 6A). This might also explain the increased size of the senescent nuclei. Thus, senescence in these irradiated cells is associated with dramatic changes in chromosome numbers. All signals of a given chromosome were included in the analyses of their nuclear positions, as shown in Figure 6B. The results are summarized in Figure 6C and show: (i) the colocalization of CSH of gene-poor, heterochromatin-rich chromosomes HSA18 and HSA4 with the INM (as indicated by LA/C) was approximately twofold higher compared with the CSH of gene-rich, heterochromatin-poor chromosomes HSA17 and HSA11 in MCF7 and U2OS cancer cells. (ii) Transitioning cells to senescence by γ-irradiation significantly reduced the fraction of CSH of gene-poor chromosomes that was colocalized with LA/C (P ≤ 0.0001). By contrast, for gene-rich chromosomes, only an insignificant decrease in CSH colocalization with the INM was observed in senescent cells compared with cycling cells. (iii) Importantly, the fraction of CSH of all chromosomes located at the INM was significantly lower in LBR(−) clones of both cell lines (MCF7 and U2OS) compared with the parental cells (P ≤ 0.0001); however, this difference was particularly significant for gene-poor chromosomes.
The location and distension of centromeric heterochromatin are related to the level of LBR and heterochromatin content of chromosomes
Thus, our results indicate that a decrease in LBR expression causes centromeric heterochromatin to detach from the nuclear membrane and relocate to the nucleoplasm. Importantly, the observation of this phenomenon in both senescent and LBR(−) cells indicated that the decrease in LBR and LB1, but not senescence per se, initiates this heterochromatin reorganization.
Distension of centromeric heterochromatin in senescent cells after γ-irradiation
FISH (fluorescence in situ hybridization) with centromeric heterochromatin probes usually provides round, compact signals in interphase cells. Such signals accounted for nearly 100% of CSH of gene-poor chromosomes HSA18 and HSA4 in cycling MCF7 and U2OS cancer cells; distension of the CSH of these chromosomes was very rare (Figure 6D). On the other hand, 30–50% of gene-rich chromosome CSH in cycling MCF7 and U2OS cells were distended, indicating that the distension of centromeric sequences is determined not only by their nuclear location but also by the heterochromatin content of the chromosome. The transition of cancer cells to senescence, accompanied by a large decrease in LBR, resulted in the significant relocation and distension of CSH in gene-poor HSA18 and HSA4 (P ≤ 0.0001) (Figure 6C,D). Statistical analysis revealed that approximately 50% of CSH of HSA18 and HSA4 had detached from the INM, relocated to the nucleoplasm, and become extended (Figures 6C,D and 7A,B). In contrast with the CSH of gene-poor chromosomes, the CSH of gene-rich chromosomes HSA17 and HSA11 were already distended in non-irradiated cycling cancer cells (Figure 6D) and induction of senescence did not lead to their marked relocation to the nucleoplasm and overall distension (Figure 6C,D). Distension of gene-poor HSA11 CSH was significant (P ≤ 0.001) only in U2OS cells after transition to senescence or stable reduction in LBR by shRNA.
Three-dimensional images of cell nuclei showing the location and distension of centromeric heterochromatin of HSA18 in (A) MCF and MCF7-LBR(−) and (B) U2OS and U2OS-LBR(−) control and senescent cells.
Distension was observed not only at the CSH located under the nuclear membrane, but also at several CSH that still colocalized with LA/C. Changes in higher order chromatin folding of these CSH were often accompanied by prolongation of the signal, occasionally forming long threads or several dots (Figures 6A and 7A,B).
Distension of centromeric heterochromatin in cycling LBR-deficient cells
One of the presumed functions of LBR is to tether heterochromatin to the nuclear membrane. We therefore wondered if a reduction in LBR levels in stable MCF7 and U2OS clones (LBR−) would lead to changes in the interphase cell nuclear topology. As shown in Figure 6C, the LBR(−) cells showed increased relocation of the CSH of gene-poor chromosomes to the nucleoplasm and distension (Figure 6D). The magnitude of CSH distension was high (P ≤ 0.0001) at gene-poor chromosomes HSA18 and HSA4 in LBR(−) clones and gene-rich HSA11 CSH in the U2OS LBR(−) clone compared with control cells. However, the CSH of gene-rich chromosome HSA17 in LBR(−) clones of both cell types was only slightly more distended compared with the control cells. Proportions of distended CSH of all chromosomes (Figure 6D) were only slightly higher after transition of LBR(−) clones to senescence after γ-irradiation compared with the control non-irradiated LBR(−) clones. These results, particularly the relocation of the CSH from the INM to the nucleoplasm accompanied by this heterochromatin distension in non-senescent LBR(−) clones, demonstrated that this process is not an exclusive property of senescence, but is provoked by the decreased level of LBR that apparently accompanies or even precedes the development of senescence.
We have shown that DNA damage induced by γ-irradiation triggers senescence in two different cancer lines: breast carcinoma MCF7 and osteosarcoma U2OS cells. Before the expression of SA-β-gal, irradiated cells reduced the expression of LBR and LB1. They also exhibited other phenotypic changes characteristic of senescence, with the exception of SAHF formation but with extensive chromatin condensation. In addition, senescent cells contained high numbers of two different types of MN: those containing euchromatin, LB1 and LA/C; and the other without lamins, containing heterochromatin and, frequently, a high level of LBR and γH2AX. This increased formation of MN suggested chromatin fragmentation after cell irradiation, probably because of reduced repair of DSBs, and their easier passage into the cytoplasm through the nuclear membrane of increased permeability. In the cytoplasm, these heterochromatic MN, that are not coated by lamins and could be therefore considered as heterochromatin fragments, are probably degraded by autophagy machinery as shown previously .
We followed the role of LBR in changes of chromatin architecture in senescent cells demonstrated by Chandra et al.  and Sadaie et al. . Both these groups showed an association between SAHF formation, gene regulation, and dynamic changes in LB1 during senescence of human diploid fibroblasts. In our experiments, cancer cells were induced to senescence by γ-irradiation and heterochromatin of these cells was not organized into SAHF, but it was extensively condensed into irregular structure (cords) densely stained by Topro 3. Thus, our finding is in accordance with a previous report  showing that SAHF formation is not a common feature of cellular senescence. SAHF were observed mainly in cultured fibroblasts promoted to senescence by oncogenes or replicative senescence where they parallel the enhanced expression of p16 INK4a. However, the level of this protein is extremely low in MCF7 and U2OS cells and its expression is not induced in senescence. It is evident from the results of Bartkova et al.  that elevated p16 INK4a may be dispensable for the induction of senescence. Moreover, it was shown that its expression varies significantly among different cell lines and it is dependent on whether p53 and Rb function in a linear manner or in parallel .
Both p53 and Rb proteins seem to respond normally to DNA damage in our experiments (Figure 3E). The level of p53 increased and that of Rb protein decreased 24 h after irradiation. We did not detect either increased expression of p16 INK4a or p15 INK4b in MCF7 and U2OS cancer cells (both parental and LBR deficient) promoted to senescence after DNA damage induced by γ-irradiation. The inhibition of the activity of cyclin-dependent kinases and thus cell cycle arrest in senescent MCF7 and U2OS cells, as well as in MCF7-LBR(−) and U2OS-LBR(−) clones, is apparently executed by the p21 protein, the level of which increased substantially 24 h after cell irradiation and gradually grew over the next 6 days. It was shown  that up-regulation of this cyclin-dependent kinase inhibitor p21 in senescent cells leads to the inhibition of pRb phosphorylation which controls cell cycle progression. The expression of the p21 protein is tightly controlled by p53 through which this protein mediates the p53-dependent cell cycle G1 phase arrest in response to a variety of stresses . The continuous increase in levels of p53 and its phosphorylated form (Ser15) started as early as 24 h PI and continued for over 6 days. The kinetics of p53 and p21 activation following irradiation was remarkably similar. It is thus probable that the activities of p53 and p21 are sufficient for maintaining cell cycle arrest in senescent MCF7, U2OS, and their LBR(−) clones after γ-irradiation even if p16 INK4a and p15 INK4b are not expressed.
How the chromatin rearrangement in senescence is related to the reduction in LB1 has not yet been determined. We attempted to answer this question by studying the role of LBR in the attachment of CSH of gene-poor and gene-rich chromosomes to the INM in cycling cancer cells and after the transition of these cells to senescence induced by γ-irradiation. We observed that a high fraction of CSH of heterochromatin-rich (gene-poor) and about 50% of heterochromatin-poor chromosomes colocalized with the INM in these cancer cells and more than half of these sequences detached from the INM and relocalized to the nucleoplasm where they were decondensed, not only in senescent cells, where LBR and LB1 were lost, but also in cells where LBR was reduced by shRNA. Colocalization of CSH with the INM also decreased significantly, by approximately two-thirds, in heterochromatin-rich and by one-third in euchromatin-rich chromosomes in cells with shRNA-down-regulated expression of LBR. These results confirmed the role of LBR as a constitutive heterochromatin tether in proliferating cancer cells, because this protein is known to contain specific regions for heterochromatin attachment, recognition of specific histone methylations and, in addition, the attachment of LB1 [25–29]. However, until recently, it was generally assumed that heterochromatin is directly tethered to the INM by lamins. Hundreds of genomic regions interacting with lamina (LAD) were identified in human fibroblasts and during differentiation [45,46]. Nevertheless, previous experiments showed that lamins do not have specific sequences to bind directly to heterochromatin and that this function is performed, with a high probability, by specific lamin-binding proteins [47–49]. LBR represents such a specific protein for LB1 and was found to be essential for tethering heterochromatin to the INM in embryonic and undifferentiated cells [29,34,35]. The binding of LBR, but not LB1, to heterochromatin also follows from the stability of LBR associated with condensed chromatin and the INM during the late stage of apoptosis, while lamin B1 is proteolysed at an early stage . The protease resistance of this LBR association, which plays a major role during apoptosis, is also likely to be important in nuclear membrane reassembly in late anaphase . It follows from these results that LBR and LB1 together form a functional unit in which LBR recognizes specific nucleotide sequences and histone modifications for heterochromatin organization, transcription repression, and attachment to the INM. LB1, which is attached to LBR in globular domain II, probably represents some kind of matrix for LBR, distributing this protein in the lamina and directing it to the chromatin regions containing specific lysine-methylated residues that should be condensed and attached to INM. Thereby, a specific chromatin structure is set and conditions are established for active transcription, including genes responsible for cell proliferation in undifferentiated, cycling cells [29,34]. Thus, the down-regulation of LBR and LB1 at the onset of senescence is necessary for the release of heterochromatin binding to the lamina, resulting in changes in chromatin architecture, as shown in the present study. These changes to chromatin architecture might give a rise to gene expression changes in senescent cells with a reduced level of LB1, as detected by Sadaie et al. . The relocation of centromeric heterochromatin containing the H3K9me3 modification from the INM to the nucleoplasm in senescent cells demonstrated in the present work is in agreement with the results of Sadaie et al. , who observed a large reduction in LB1 in H3K9me3-enriched chromatin regions. These authors also observed small portions of genome, particularly in H3K27me3-enriched regions, which gained LB1 binding during senescence which suggests regulated redistribution of reduced LB1 to specifically modified heterochromatin. As there are unknown specific reader sequences of LB1 for the recognition of lysine-methylated residues of chromatin, it is highly probable that this function is executed by LBR, the level of which was not followed by the authors; however, they anticipated the existence of such a mediator of LB1. Our results showing the detachment of CSH of chromosomes with the prevailing amount of heterochromatin (mainly of chromosome 18) from INM in cells with reduced expression of LBR and LB1 are also in consensus with the results of Malhas et al. , which showed that an absence of LB1 or its full length protein resulted in relocation of chromosome 18 from the nuclear periphery to the nuclear center in mouse embryonic fibroblasts, followed by decondensation of this chromosome and changes in expression of some of its genes. No changes were observed in chromosome 19, located in the center of these cells' nuclei. Even if the authors did not follow changes in LBR level, it could be supposed, on the basis of our results, that in the absence of LB1 or in the presence of its incomplete structure, the heterochromatin tether, executed in these embryonic cells by LBR , cannot be performed, resulting in the exchange of a normal location of chromosome 18 from the nuclear membrane to the nuclear center. It follows from our results that both LBR and LB1 are needed for tethering of constitutive heterochromatin to the INM and maintaining lamina integrity in proliferating cells. Down-regulation of LBR leads to the down-regulation of LB1 showing the dependence of LB1 expression on LBR. However, down-regulation of both these proteins by LBR shRNA did not induce senescence. Clones of MCF7 and U2OS with reduced expression of these proteins showed slower proliferation compared with the parental cells, formed higher numbers of MN showing higher permeability of the nuclear membrane, and were transferred to senescence by γ-irradiation similarly to their parental cells. Our results show that the down-regulation of LBR and LB1 is necessary for cell transition to senescence. An increased activation of autophagy/lysosomal catabolic mechanisms processing proteins of the nuclear membrane and chromatin fragments extruded into the cytoplasm that was observed during cellular senescence [18,21] could play an important role in more profound changes of cellular functions after down-regulation of LBR and LB1. Current findings show that cellular senescence is triggered by various stresses, inducing DNA damage [e.g. by γ-irradiation shown in the present study or by replication stress shown in our yet unpublished results (Lukasova et al. in preparation)], telomere shortening and activation of oncogenes  that could also induce apoptosis or cancer. We propose, on the basis of the presented results, that the main condition preventing apoptosis or development of cancer upon activation of these stresses is a down-regulation of LBR and LB1 which is indispensable for the rearrangement of chromatin architecture. However, the principal impulse triggering the down-regulation of these proteins is not yet known.
centromere-specific satellite heterochromatin
fluorescence in situ hybridization
inner nuclear membrane
- LA/C or LAC
lamin B receptor
clone of MCF7 (or U2OS) cells with LBR expression down-regulated by shRNA
senescence-associated heterochromatin foci
Emilie Lukášová conceived and designed the project. Aleš Kovařík, Emilie Lukášová and Alena Bačíková performed experiments, analyzed and interpreted data. Martin Falk performed statistical analysis and construction of images. Emilie Lukášová wrote the paper with a help of Aleš Kovařík and Martin Falk. Stanislav Kozubek participated on data interpretation.
The study was supported by the Czech Science Foundation [project P302/12/G157 and P501/12/G090] and MYES CR (Project of the Czech Plenipotentiary 2016, the 3+3 Project for 2016-18.
We thank the Edanz editing team for English editing of our manuscript, and Professor Marketa Hermanova, PhD, Institute of Pathology – University Hospital St. Anna in Brno for scanning the senescent cells.
The authors declare that there are no competing interests associated with the manuscript.