Introduction

The nuclear envelope (NE) encloses genomic DNA to protect it from a variety of insults and plays pivotal roles in regulating genomic function [1, 2]. Previous studies demonstrated that the NE is damaged upon various types of stresses, leading to its dysfunction [3,4,5,6,7]. Such stresses are referred to as “NE stress” [8,9,10,11]. Mechanical NE stress frequently results in the rupture of the NE [3, 4]. Cell migration through confined environments is one of the major causes of NE rupture [3, 4]. This can include both physiological and pathological events such as the radial migration of neural progenitors during brain development and the invasion of tumor cells into surrounding tissues, respectively [3, 4, 12]. Accumulating evidence has revealed that NE rupture is involved in a broad spectrum of diseases ranging from cancer to neurodegenerative diseases. Regarding the pathology of cancer, NE rupture has been proposed to trigger genome instability and increased cellular invasion through DNA damage [13, 14]. Mechanistically, studies have suggested that the mislocalization of cytoplasmic DNase and/or nuclear DNA repair factors is responsible for such DNA damage [14, 12]. These findings imply that the consequences of NE stress might be context dependent. Thus, we investigated whether mechanical NE stress causes apoptosis in our model. To achieve this, U251MG, U87MG, and NP5 were plated on a Transwell, and apoptosis was detected by immunostaining using anti-cleaved caspase-3 antibodies. Substantial cleaved caspase-3 signals were observed at the bottom side of the Transwell for U251MG and NP5, whereas for U87MG they were hardly detected (Fig. 7A, B). These findings suggest that mechanical NE stress results in the apoptotic cell death of U251MG and NP5, presumably due to inefficient repair of ruptured NE. To test whether p21 plays pivotal roles in the suppression of NE stress-mediated apoptosis, U87MG cells were subjected to a Transwell assay with and without depletion of p21. Depletion of p21 resulted in a significant increase of the presence of the clusters of cleaved-caspase-3 (Fig. 7C, D). Thus, it is suggested that p21 is indispensable for the suppression of apoptosis induced by NE stress in U87MG cells.

Fig. 7: Ineffective NE repair causes apoptotic cell death after constricted migration.
figure 7

A Immunofluorescence staining analysis of cleaved caspase-3 (green) of U251MG, U87MG, and NP5 at the bottom side of a Transwell. Overlay indicates merged images with DNA (blue). Scale bars: 10 μm. B Quantification of cleaved caspase-3 clusters relative to total cell number in the cells indicated in (A). Bars and error bars represent the mean values and standard deviations from three independent experiments. Statistical significance of the differences was determined using Fisher’s exact test. ***p < 0.005. Total cell numbers for each cell line from three independent experiments: n = 1045, U251MG; n = 701, U87MG; n = 357, NP5. C Immunofluorescence staining analysis of cleaved caspase-3 (green) of U87MG cells with (sip21) or without (siCon) RNAi of p21 at the bottom side of a Transwell. Overlays represent merged images with DNA (blue). Scale bars: 5 μm. D Quantification of cleaved caspase-3 clusters relative to total cell number in the cells indicated in (C). Bars and error bars represent the mean values and standard deviations from three independent experiments. Statistical significance of the differences was determined using Fisher’s exact test. *p < 0.05. Total cell numbers for each condition from three independent experiments: n = 540, siCon; n = 1055, sip21.

Discussion

While some of the molecular mechanisms of NE stress response have been clarified, how such processes are affected by the types of cells and/or the cellular environment has not been elucidated. In the present study, we demonstrated that the response to mechanical NE stress varies among cancer cell lines. Furthermore, our data revealed the impact of the cell cycle on sensitivity to mechanical NE stress in GBM. Mechanistically one of the major CDK inhibitors, p21, is indispensable for the resistance to mechanical NE stress and suppresses the consequences of NE stress.

In higher eukaryotes, the NE is completely disassembled at the early onset of mitosis to allow access to the mitotic spindle to the chromosomes [22]. Thus, disassembly and subsequent reassembly of the NE are essential for the progression of the cell cycle. Prior to the disassembly of the NE, many NE components, including BAF and LEM domain proteins, are dissociated from the NE [22]. The change of their localization is mainly regulated by cell-cycle-dependent phosphorylation through several kinases including CDK1 [22, 24, 25, 32]. After the segregation of chromosomes, NE components are recruited to the surface of the chromosomes to establish functional NE. In contrast to the disassembly of the NE, reassembly requires the dephosphorylation of many NE proteins, which is mediated by the inactivation of mitotic kinases [22]. Currently, BAF is known as the earliest accumulating factor to the chromosome among the proteins that are dissociated from the NE prior to NE disassembly [22, 23, 33]. BAF targets LEM domain proteins, leading to further recruitment of ESCRT III complex to coordinate reseal the NE and remove the spindle from chromosomes [22, 34]. Some studies have proposed that the repair of ruptured NE involves a process similar to NE reassembly, featuring sequential targeting of BAF, LEM domain proteins, and ESCRT III to the ruptured site [16, 17, 23]. Thus, it is highly possible that the regulatory mechanisms are also shared between the repair and reassembly of the NE. In this scenario, phosphorylation of the NE repair factors related to the disassembly of the NE will perturb the efficiency of NE repair due to the loss of their capacity to accumulate at the ruptured NE. This can explain why G2-phase of cells are more sensitive to NE stress even if the expression of p21 is absent in the entire population of U251MG cells. In this regard, p21 can restore effective NE repair by inhibiting the phosphorylation of NE repair factors. Furthermore, because p21 is depleted in the S phase in U87MG cells, sensitivity to mechanical NE stress is possibly higher at this stage of cell cycle in this cell line as well as in other cancer cells with the same pattern of p21 expression. Dysregulation of the cell cycle is frequently observed in many types of cancer, which potentially impairs the efficient repair of ruptured NE [35]. Our data further support this idea and reveal the link between the cell cycle and NE stress response.

It is possible that cell-cycle-dependent perturbation of NE repair occurs not only in GBM but also in other types of cancers, although we have not observed severe consequences of mechanical NE stress in those cancer cell lines (Fig. 1). Additional investigations will clarify whether the unknown cell-cycle-independent NE repair pathway plays pivotal roles to maintain NE integrity upon mechanical NE stress in those cells. Moreover, further studies are also required to elucidate the significance of apoptosis caused by NE stress in NE stress sensitive cells. Regarding the malignancy of GBM, it is possible that G1-phase cells are responsible for the invasion of the GBM in vivo due to their resistance to mechanical NE stress.

In summary, this study revealed that (1) the efficiency of repair of ruptured NE is regulated in a cell-cycle-dependent manner and (2) impaired NE repair caused by dysregulation of cell cycle triggers severe consequences of NE stress in GBM. These novel insights indicate that further detailed investigation of the molecular mechanisms behind the regulation of NE repair should make significant contributions to the development of an innovative therapeutic strategy of GBM.

Materials and methods

Cell culture, plasmids, and stable cell lines

U251MG, HT1080, DU145, A549, and U87MG cells were maintained in Eagle’s Minimum Essential Medium (FUJIFILM Wako Pure Chemical Corporation, Osaka, Japan) supplemented with 10% (v/v) heat-inactivated fetal bovine serum (FBS), 1 mM sodium pyruvate, 1X non-essential amino acids, and penicillin/streptomycin at 37 °C in a 5% CO2, 95% humidified air atmosphere. MDA-MB468 cells were maintained in Leibovitz’s L-15 Medium (FUJIFILM Wako Pure Chemical Corporation) supplemented with 10% FBS and penicillin/streptomycin at 37 °C in 100% air atmosphere. NP5 cells were maintained in Dulbecco’s Modified Eagle Medium (FUJIFILM Wako Pure Chemical Corporation) supplemented with 10% FBS and penicillin/ streptomycin at 37 °C in 5% CO2, 95% humidified air atmosphere. Stable cell lines were established using either transposon or lentivirus. Plasmids for Slee**beauty based transposon vectors were obtained from addgene. Donor plasmids (pSBbi-Bla: #60526; pSBTet-Pur: # 60507) were transfected together with the expression plasmid for a mutant form of transposase, SB100X (pCMV(CAT)T7-SB100: #34879) using Everyday Transfection (EZ Biosystems, College Park, MD, USA) according to manufacturer’s protocol [36, 37]. A NLS derived from SV40 large T antigen was fused with N-terminus of tdTomato by PCR and resulting fragment was cloned into pSBbi-Bla vector (amplified with PCR) using In-Fusion system (TaKaRa Bio, Kusatshu, Shiga, Japan). cDNA of human p21 was obtained by reverse transcription using ReverTra Ace (TOYOBO, Osaka, Japan) and RNA purified from U87MG. PCR was performed to clone p21 into pmVenus-C1 using BsrGI/EcoRI sites [9]. Then, resulting mVenus-p21 was amplified with PCR and cloned into pSBTet-Pur (digested with SfiI) using In-Fusion system. The stable cell lines expressing mScarlet-BAF were generated by transposon-based vectors indicated above. cDNA of BAF is amplified by PCR using EGFP-BAF plasmid (addgene number: 52962) and cloned into pSBbi-Bla together with mScarlet-I (a gift from Dorus Gadella, addgene number: 98831) [38]. The primers used for construction are listed in Supplementary Table 1. Integrated cells were selected using blasticydin (8 μg/ml) or puromycin (2 μg/ml). The plasmid for FUCCI was obtained from RIKEN BRC (#RDB15454) [30]. To introduce FUCCI reporter set, pCSII lentivirus vector carrying mCherry-hCdt1(1/100)Cy(-) and mTurquoise-hGem(1/110) was transfected into HEK293T cells together with LP1, LP2, and VSV-G plasmids. Resulting lentivirus particles were concentrated using Lenti-X Concentrator (TaKaRa Bio) according to manufacturer’s protocol. Polybrene (8 μg/ml) was supplemented to the medium when cells were infected with lentiviruses.

Constricted migration assay using Transwell

Constricted migration assay was performed using Transwell with a 3 μm pore size (Corning Incorporated, Corning, NY, USA) as described previously with minor modifications [9]. 1 × 105 cells were plated on a Transwell and cultured for 14–16 h, followed by immunofluorescence staining. For immunofluorescence staining of cleaved-caspase-3, cell were cultured for 40 h.

EdU labeling and click reaction

To identify S-phase cells, EdU (Jena Bioscience, Jena, Thuringia, Germany) was added into culture medium at 25 μM and incubated for 30 min. Incorporated EdU was visualized by a fluorescent dye through a Cu(I)-catalyzed [3 + 2] cycloaddition (click) reaction, by which terminal alkyne group of EdU is covalently attached with azide group of a fluorescent dye, as described previously with minor modifications [39, 40]. Briefly, cells were fixed with 4% paraformaldehyde (PFA) for 15 min and permeabilized with 0.5% Triton X-100 for 3 min. After three times PBS washes, cells were incubated with the reaction mixture (2 μM picolyl azide Alexa-555 (Jena Bioscience), 20 mM Sodium ascorbate (FUJIFILM Wako Pure Chemical Corporation), 2 mM 2-(4-((bis((1-(tert-butyl)-1H-1,2,3-triazol-4-yl)methyl)amino)methyl)-1H-1,2,3-triazol-1-yl)acetic acid (Jena Bioscience), 1 mM CuSO4 (FUJIFILM Wako Pure Chemical Corporation) diluted in PBS) for 30 min at 37 °C. After three times PBS wash, cells were subjected to the immunostaining if necessary.

Immunofluorescence staining and nuclear circularity

The cells were grown on coverslips (Matsunami Glass, Osaka, Japan) or Transwell and fixed in 4% PFA for 15 min. After three times wash with PBS, the cells were permeabilized in 0.5% Triton‐X 100 for 3 min followed by blocking with 10% normal goat serum for 60 min. The following antibodies were used as the primary antibodies: anti‐lamin B1 (1:1000; ab16048, Abcam, Cambridge, UK), anti-γH2A.X (1:1000; 05-636, Merk Millipore, Burlington, MA, USA), anti-cyclin B1 (1:200; SC-245, Santa Cruz Biotechnology, Dallas, TX, USA) anti-p21 (1:500; #2947, Cell Signaling Technology, Danvers, MA, USA), anti-cleaved caspase-3 (1:500; #9661, Cell Signaling Technology). Incubation with primary antibodies was performed at 4 °C overnight or 2 h at R.T. For secondary antibodies, goat anti-mouse or anti-rabbit IgG F(ab′)2 fragments conjugated with Alexa 488 or 568 were used (Thermo Fisher Scientific, Waltham, MA, USA). Confocal fluorescent images were acquired using FV1000D (Evident, Tokyo, Japan), FV3000 (Evident, Tokyo, Japan), or Stellaris5 (Leica Microsystems, Wetzlar, Germany). All images were processed through imageJ (National Institutes of Health, Rockville, MD, USA) [41]. The prominent signal of γH2A.X is defined manually with same setting of contrast among a series of the experiments. Obtained images of DNA staining using DAPI were processed to measure nuclear circularity as bellow using imageJ. All the images were subjected to “Gaussian Bluer (sigma 1)” and then Z-stacked with maximum intensity. The binary images were generated using “Threshold (Li, auto)” and subjected to “Fill holes” filter. Then, the nuclear circularity was measured using “Find particles”. Overlapped nuclei were manually excluded.

Laser irradiation

U251MG and U87MG cells expressing NLS-tdTomato or mScarlet-BAF were plated on the glass bottom dish (Matsunami glass, Osaka, Japan) coated with rat collagen I (Thermo Scientific). Laser irradiation and time lapse imaging was performed using FV1000D. A circular spot with 0.6 μm diameter at the periphery of the nucleus was irradiated with a 405 nm laser at 100% power for 6 s using a SIM light pathway. DNA was stained with SPY650 DNA probe (SPIROCHROME, Thurgau, Switzerland) and visualized with a 635 nm laser. Time lapse imaging was performed with 3 min intervals for 57 min. Obtained images of NLS-tdTomato were processed to measure the ratio of NLS-tdTomato signal in cytoplasm to the nucleus as bellow using imageJ. All the images were subjected to “Subtract Background (Rolling ball radius: 50 pixels, Sliding paraboloid, Disable smoothing), “Gaussian Bluer (sigma 1)”, and then Z-stacked with maximum intensity. Some regions of the nucleus and cytoplasm were manually selected and the mean signal intensity were measured. The signal intensity of cytoplasmic NLS-tdTomato was divided by that of the nucleus, and its highest values during the time course was set as 1.0 for each cell. For the immunostaining of γH2A.X after laser irradiation, the cells were cultured additional 16 h and subjected to immunostaining. All the regions that possibly include irradiated cells were visualized with tile scan using Stellaris5.

Protein preparation and western blotting

Proteins were extracted from the indicated cells in RIPA buffer (50 mM Tris-HCl (pH 7.5), 150 mM NaCl, 0.1% SDS, 0.5% Sodium deoxycholate, 1% Triton X-100) supplemented with Protease inhibitor cocktail Set V/Phosphatase inhibitor cocktail Set (FUJIFILM Wako Pure Chemical Corporation). The lysates were incubated for 30 min on ice. After centrifugation at 15,000xg for 5 min, the protein concentrations of the supernatants were determined using a bicinchoninic acid assay kit (Thermo Fisher Scientific). Equal amounts of proteins were used for SDS-PAGE. For immunoblotting, the following antibodies were used: anti-cyclin B1 (1:200; SC-245, Santa Cruz Biotechnology, Dallas, TX USA), anti-cyclin E (1:200; SC-247, Santa Cruz), anti-p21 (1:2000; #2947, Cell Signaling Technology), anti‐lamin A/C (1:2000; MABT1341, Merk Millipore), anti-MAN1 (1:10000; TA500794, OriGene, Rockville, MD USA) anti‐β‐actin (1:5000; MAB1501, Merk Millipore).

Flow cytometry

The cells were trypsinized, washed with PBS once, and fixed with 70% of EtOH for 30 min at 4 °C. After PBS wash, the cells were stained with PI solution (0.15 % Triton X-100, 2μg/ml PI, 50μg/ml RNaseA) at R.T. for 30 min. Cell cycle profile was obtained using FlowJo.

RNA interference

For knockdown of p21, small interfering (si)RNA targeted to human p21 (AM51331) and negative control (AM4611) were purchased from Thermo Fisher Scientific. siRNA were transfected using Lipofectamine RNAiMax (Thermo Fisher Scientific) as described in the manufacture’s manual. The depletion of p21 was confirmed by western blot 48 h after transfection. Transfection was performed on the top side of the Transwell without the medium at the bottom side. After 12 h Transfection, the medium was added to the bottom side, then the cells were cultured additional 36 h before immunostaining.

Statistical analysis

Microsoft Excel and R (version 4.0.2) were used via RStudio (version 1.3.1073) for statistical analysis [42, 43].