Inactivation of Mad2B Enhances Apoptosis in Human Cervical Cancer Cell Line upon Cisplatin-Induced DNA Damage

Abstract

Mad2B (Mad2L2), the human homolog of the yeast Rev7 protein, is a regulatory subunit of DNA polymerase ζ that shares sequence similarity with the mitotic checkpoint protein Mad2A. Previous studies on Mad2B have concluded that it is a mitotic checkpoint protein that functions by inhibiting the anaphase-promoting complex/cyclosome (APC/C). Here, we demonstrate that Mad2B is activated in response to cisplatin-induced DNA damage. Mad2B co-localizes at nuclear foci with DNA damage markers, such as proliferating cell nuclear antigen and gamma histone H2AX (γ-H2AX), following cisplatin-induced DNA damage. However, unlike Mad2A, the binding of Mad2B to Cdc20 does not inhibit the activity of APC/C in vitro. In contrast to Mad2A, Mad2B does not localize to kinetochores or binds to Cdc20 in spindle assembly checkpoint-activated cells. Loss of the Mad2B protein leads to damaged nuclei following cisplatin-induced DNA damage. Mad2B/Rev7 depletion causes the accumulation of damaged nuclei, thereby accelerating apoptosis in human cancer cells in response to cisplatin-induced DNA damage. Therefore, our results suggest that Mad2B may be a critical modulator of DNA damage response.

INTRODUCTION

The Y-family of DNA polymerases (Pol η, Pol κ, Pol τ, and Rev1) carries out translesional DNA synthesis (TLS), enabling cells with damaged DNA to continue DNA replication (Prakash et al., 2005). Another TLS enzyme, DNA polymerase zeta (Pol ζ), is composed of two subunits (Rev3 and Rev7), and plays an important role in translesional bypass replication in most eukaryotes, including humans (Gan et al., 2008). The Rev3 subunit forms the catalytic core of DNA Pol ζ (Nelson et al., 1996); however, the function of the Rev7 subunit remains elusive. A comparison of the amino acid sequences of Rev7 and Mad2B indicates that they are identical proteins (Murakumo et al., 2000). Mad2B was initially isolated in a screening for human mitotic checkpoint genes (Cahill et al., 1999). Mad2B (also known as Mad2-Like 2 or Mad2L2) displays 43% amino acid sequence identity to the mitotic checkpoint protein Mad2 (Cahill et al., 1999), suggesting that Mad2B also functions as a mitotic checkpoint protein. The mitotic checkpoint (also known as the spindle assembly checkpoint or SAC) is a signaling mechanism that suppresses anaphase onset in the presence of unattached kinetochores or anti-microtubule agents, such as nocodazole (Primorac and Musacchio, 2013), thereby preventing unequal segregation of chromosomes. The main effector of the SAC is the mitotic checkpoint complex (MCC) composed of the mitotic checkpoint proteins BubR1 (or Mad3), Bub3, and Mad2, which together bind to Cdc20, the co-activator of the anaphase-promoting complex or cyclosome (APC/C). Despite increasing evidence that Mad2B plays an important role in DNA damage response (DDR) in mammalian cells, the role of Mad2B in the mammalian mitotic checkpoint is currently elusive (Okada et al., 2005; Cheung et al., 2006; Boersma et al., 2015; Xu et al., 2015; Clairmont and D’Andrea, 2021)

The mechanisms of DNA damage checkpoints in mammalian cells have been extensively characterized, and their function is critical for the maintenance of genomic integrity (Harper and Elledge, 2007). DNA damage also initiates a DDR that ultimately results in the recruitment of repair proteins to sites of DNA damage (Bennett and Harper, 2008). The ubiquitylation of proliferating cell nuclear antigen (PCNA) by E3 ubiquitin ligases such as Rad18 is critical for the recruitment of TLS polymerases to sites of DNA damage, thereby replacing replicative polymerases (Leung et al., 2019). Mad2B/Rev7 function has been implicated in DNA damage checkpoint and SAC; however, the molecular details remain elusive. Additionally, we used cisplatin as a DNA-damaging agent in human cancer cells. Cisplatin, cis-diamminedichloroplatinum (II), is a well-known chemotherapeutic agent used to treat a variety of human cancers, and works in part by binding to DNA, inhibiting its replication, and causing DNA damage (Dasari and Tchounwou, 2014).

In this study, we aimed to elucidate the role of Mad2B/Rev7 in the cell cycle of human cancer cells and to evaluate whether it is a key regulator of the cellular response to cisplatin-induced DNA damage.

MATERIALS AND METHODS

Cell culture and preparation of cell extracts

Human cervical carcinoma cells (HeLa) were obtained from BioWhittaker Europe (Lonza, Slough, UK). To obtain mitotically-arrested cells, an asynchronous population of HeLa cells was treated (3 µM) for various times (0-30 h). Mitotic cells were collected by mechanical shake-off, washed in Dulbecco’s phosphate-buffered saline (DPBS) and lysed in Radio-immunoprecipitation buffer (RIPA) (Thermo Scientific, Rockford, IL, USA) supplemented with protease and phosphate inhibitor cocktail (Thermo Scientific). Cell lysates were centrifuged at 14,000 g for 10 min at 4°C and the protein content of the supernatant determined using the Coomassie Protein Assay Reagent (Thermo Scientific) before normalisation and solubilisation in x2 SDS-PAGE sample buffer. Soluble and chromatin protein fractions were prepared as described in previous report (Gillotin, 2018).

DNA damage

To induce double strand DNA damage exponentially growing HeLa cells were treated with cisplatin (50 µM) for varying times.

Plasmids

The genes for human Mad2A and Mad2B were cloned by PCR amplification from a human HeLa cell cDNA library (Clontech, San Jose, CA, USA) using sequence-specific primers. One set of primers was also designed to incorporate an HA-epitope tag at the 5’ terminus of both the Mad2A and Mad2B genes. The non-epitope tagged genes were subcloned into pGEX-4T-1 (Amersham-Pharmacia, Buckinghamshire, UK) and the HA-epitope tagged Mad2A and Mad2B genes were subcloned into a mammalian expression vector (pCMV5). The pEF-4myc-Cdc20 expression vector was generated by cloning full length human Cdc20 downstream of 4 myc tags under an elongation factor 1 promoter.

Antibodies

The following antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA): monoclonal anti-HA, rabbit polyclonal anti-HA and polyclonal anti-Cdc20. Rabbit polyclonal antibodies to Chk1, PChk1 (Ser317), Cdc27, PHistone-H2AX (Ser139), Histone H3 antibodies and mouse monoclonal FLAG antibody were purchased from Cell Signaling Technology (Beverly, MA, USA). The following antibodies were purchased from Sigma-Aldrich (St Louis, MO, USA): monoclonal anti-γ-tubulin, polyclonal α-tubulin, horseradish peroxidase (HRP)-conjugated goat anti-mouse, HRP-conjugated goat anti-rabbit. A human anti-centromere antibody was purchased from Europa Bioproducts Ltd (Cambridge, UK). Monoclonal antibodies to Mad2B, Mad2A, PCNA and c-myc were purchased from BD Transduction Laboratories (Heidelberg, Germany). The following antibodies were purchased from Invitrogen (Carlsbad, CA, USA): Alexa Fluor 488 goat anti-mouse IgG, Alexa Fluor 488 goat anti-rabbit IgG, Alexa Fluor 594 rabbit anti-mouse IgG and Alexa Fluor 594 goat anti-human IgG.

Transfection of HeLa cells

HeLa cells were transfected using Fugene 6 (Roche Diagnostics Ltd, Mannheim, Germany), according to the manufacturer’s instructions.

RNAi

The Mad2B oligonucleotides and non-targeting, control siRNA oligonucleotide was purchased from Dharmacon Research Inc (Lafayette, CO, USA). The sequence of the sense strand of the siRNA duplexes were as follows: Mad2B, 5’-CCAAAGUUGAGGUCUUGUCUU-3’, Control, 5’-UAGCGA CUAAACACAUCAA-3’. The siRNA oligonucleotides were dissolved in RNAse-free buffer (20 mM KCl, 6 mM Hepes, 0.2 mM MgCl₂, pH 7.5) at a concentration of 20 µM. Cells were transfected with the siRNA oligonucleotides using Interferin transfection reagent (Polyplus, New York, NY, USA) according to the manufacturers protocol.

Immunoprecipitation and Western blot analysis

Immunoprecipitations were performed as described previously (Deacon et al., 2003). In some experiments the immunoprecipitates were not heated in order to minimise degradation of the IgG. The immunoprecipitated proteins and the cell extracts were resolved by SDS-PAGE and electroblotted onto Hybond-C nitrocellulose membrane (Amersham-Pharmacia, Bucks, UK) using a Hoeffer semi-dry blotting apparatus (Amersham-Pharmacia, Freiburg, Germany). Immunoreactive proteins were visualised using enhanced chemiluminescence (ECL) according to the manufacturer’s instructions (Geneflow, Lichfield, UK).

Purification of recombinant glutathione-S-transferase (GST) fusion proteins

Recombinant GST-Mad2A and GST-Mad2B were prepared as described previously (Deacon and Blank, 1999). FPLC purification was performed using an ÄKTA system (GE Healthcare, Buckinghamshire, UK). GST-tagged Mad2A and Mad2B proteins were purified by passing them twice through a glutathione-agarose column (Sigma-Aldrich). GST-tagged Mad2A and Mad2B proteins were incubated with TEV protease (1500 U) provided by the Protein Expression Laboratory (PROTEX) (University of Leicester, Leicester, UK) to cleave the GST-tag. To further remove minor contaminating proteins from the Mad2A and Mad2B fractions both Ion exchange and gel filtration on a Superdex 75 (GE Healthcare, Cambridge, UK) chromatography were performed.

In vitro protein binding assays

To examine the interaction between Cdc20 and Mad2A/Mad2B interphase HeLa cells were lysed in buffer A and Cdc20 immunoprecipitated from the cell lysates. The immunoprecipitates were incubated (4ºC for 1 h) with purified, recombinant Mad2A, Mad2B or GST at varying concentrations (0.4-8 nM). The Cdc20 immunoprecipitates were washed 3 times in cold (4ºC) Dulbecco’s PBS and analysed by immunoblotting with antibodies to Mad2A, Mad2B, GST and Cdc20.

Flow cytometry

For analysis of the cell cycle, cells were harvested and collected by centrifugation at 230 xg for 5 min and washed 3 times with Dulbecco’s PBS. After washing, the cells were fixed in 70% ice-cold ethanol and stored at –20ºC (minimum period 10 min) until ready for staining. Fixed cells were washed with Dulbecco’s PBS to remove ethanol completely and resuspended in 1 mL stain solution (PBS supplemented with 500 μg/mL RNase (Sigma-Aldrich, San Diego, CA, USA) and 20 μg/mL propidium iodide (PI) (Sigma-Aldrich). Cells were incubated in the dark at room temperature for a minimum of 15 min and maximum of 6 h before analysis. The cellular DNA content was then analyzed by flow cytometry (FACScan, Becton Dickinson, Oxford, UK).

Immunofluorescence microscopy

Cells were fixed and processed for immunofluorescence microscopy as described previously (Deacon et al., 2003). Where indicated cells were pre-extracted with 0.1% v/v Triton X-100 in Dulbecco’s PBS for 2-5 min prior to fixation and immunostaining. The cell images were obtained with a TE300 inverted fluorescence microscope (Nikon, Tokyo, Japan) using Volocity imaging software (PerkinElmer Inc, Waltham, MA, USA).

In vitro ubiquitylation assay

The in vitro ubiquitylation assays were performed as described previously (Lara-Gonzalez et al., 2011). HeLa cells were treated with 0.2 μg/mL nocodazole for 16 h and the mitotically-arrested HeLa cells collected by “shake off” and lysed with lysis buffer (100 mM NaCl, 10 mM Tris-HCl pH 7.4, 0.1% v/v Triton X-100, 1 mM EDTA, 1 mM EGTA, 20 mM β-glycerophosphate, 1 mM DTT, 10 mM NaF, 0.2 mM PMSF and protease inhibitors). BubR1 was immunodepleted from lysates using an anti-BubR1 antibody bound to protein G sepharose beads (GE Healthcare), for 1 h at 4°C. APC/C was then isolated from the supernatant using anti-Cdc27 antibodies coupled to protein A beads (GE Healthcare) and washed with cold (4°C) PBS. Beads were incubated with recombinant proteins, each at a final concentration of 1.5 μM in Dulbecco’s PBS containing 10% glycerol and 4 mg/mL BSA, at room temperature for 45 min. After washing, the beads were resuspended in 15 μL of a reaction mixture (100 mM NaCl, 10 mM Tris-HCl pH 7.5, 100 nM E1, 1 μM UbcH10, 1.25 mg/mL ubiquitin, 2 mM ATP, 0.1 mg/mL creatine phosphate, 7.5 mM creatine phosphate and 2 mM MgCl₂). The E1 enzyme, UbcH10 (E2) and Ubiquitin were purchased from Boston Biochem (Boston, MA, USA). 0.2 μM of Myc-Cyclin B1 N90 was used as substrates of APC/C. Ubiquitylation reactions were incubated in a Thermomixer (Eppendorf, Hamburg, Germany) at 37ºC for 20 min, stopped by addition of SDS-PAGE sample buffer, then resolved on a 4-12% NuPAGE Bis-Tris gel (Invitrogen), transferred to a PVDF membrane (Millipore, Bedford, MA, USA) and blotted with an anti-Myc antibody.

Reagents

All other reagents were of analytical grade and obtained from either Sigma-Aldrich or Thermo Scientific.

Statistical analysis

All data are presented as the mean ± SD. The n-value represents the number of independent samples used in the experiments. The significance for all pairwise comparisons of interest was assessed using the two-tailed Student’s t-test, and p<0.05 was considered statistically significant. All experiments were independently performed with at least three different samples. GraphPad Prism 4 software (GraphPad Software, La Jolla, CA, USA) was used for statistical analysis.

RESULTS

Mad2B does not function like the SAC protein Mad2A

To determine whether Mad2B functions as a mitotic checkpoint protein, we examined its subcellular localization in response to nocodazole-induced SAC activation. As Mad2A and Mad2B antibodies were unable to detect the native protein in fixed cells, we tagged Mad2A and Mad2B with an HA tag at their N-termini. HeLa cells were transfected with either HA-Mad2A or HA-Mad2B for 48 h and subsequently treated with nocodazole for 18 h to activate the SAC. Cells were pre-extracted with Triton X-100 buffer before fixation at the indicated times to analyze the distribution of Mad2A, Mad2B, and centromeres using immunofluorescence microscopy. The results showed that HA-Mad2A was uniformly distributed in the nucleus of interphase cells (0 h) and was enriched on kinetochores (as assessed by staining with an anti-centromere antibody) following the addition of nocodazole (18 h) (Fig. 1A). In contrast, HA-Mad2B had a slightly granular distribution in the nuclei of interphase cells (Fig. 1B, 0 h). Following nocodazole addition, HA-Mad2B did not localize to the kinetochores but was homogeneously distributed throughout the cytoplasm in mitotically arrested cells (Fig. 1B, 18 h). To confirm the differential localization of Mad2A and Mad2B in a single SAC-activated cell line, we generated N-terminus-FLAG-tagged Mad2A. HeLa cells were co-transfected with Flag-Mad2A and HA-Mad2B for 48 h. Western blot analysis of the transfected cells confirmed the co-expression of FLAG-Mad2A and HA-Mad2B proteins (Fig. 1C). Parallel samples of the co-transfected cells were treated with nocodazole for 18 h. In nocodazole-arrested cells, HA-Mad2B was present almost exclusively in the cytoplasm, whereas FLAG-Mad2A formed distinct foci on the chromatin characteristic of kinetochore localization (Fig. 1D).

Fig. 1.

Mad2B does not localize to the kinetochores following activation of the mitotic checkpoint. HeLa cells were transfected with HA-Mad2A (A) or HA-Mad2B (B). After 48 h, cells were treated with nocodazole (3 μM) for 0-18 h and extracted. HA-Mad2A (green), centromere (red), DNA stained with Hoechst 33342 (blue). Scale bar, 10 μm. (C) HeLa cells were co-transfected with FLAG-Mad2A and HA-Mad2B. Control cells were mock transfected for 48 h, after which, the cells were lysed, and the cell extracts immunoblotted with either an HA- or FLAG-epitope tag antibody. (D) HeLa cells were co-transfected with FLAG-Mad2A and HA-Mad2B. After 48 h, cells were treated with nocodazole (3 μM) for 18 h and the mitotic cells were collected by shake-off. The mitotic cells were extracted and processed for immunofluorescence microscopy. HA-Mad2B (green), FLAG-Mad2A (red), DNA (blue). Scale bar, 10 μm.

To ascertain whether Mad2B behaves as a SAC protein at the molecular level, we examined whether Mad2B associates with Cdc20 in nocodazole-arrested mitotic cells. Endogenous Cdc20 was immunoprecipitated from cell lysates of both exponentially growing and mitotically arrested HeLa cells (Fig. 2A, upper panel). Western blotting of the immunoprecipitants with either a specific Mad2A or a specific Mad2B antibody indicated that endogenous Mad2B did not form a complex with Cdc20 before or after activation of the SAC (Fig. 2A, lower panel). In contrast, endogenous Mad2A co-immunoprecipitated with Cdc20 in nocodazole-arrested cells (Fig. 2A, middle panel). To identify the functional role of Mad2B in the SAC, which was independent of Cdc20-binding, we used an siRNA approach to deplete endogenous Mad2B protein. Western blot analysis of extracts prepared from cells transfected with Mad2B siRNA (5-100 nM) resulted in effective and specific depletion of Mad2B without affecting Mad2A protein levels (Fig. 2B). Depletion of the Mad2B protein did not affect the function of the Mad2A-dependent SAC, as FACS analysis of Mad2B-depleted cells indicated that nocodazole induced a stable G2/M-phase block (Fig. 2C). Therefore, our results suggest that the localization, interaction with Cdc20, and function of Mad2B in SAC-activated cells are considerably different from those of Mad2A.

Fig. 2.

Mad2B does not bind Cdc20 in SAC-activated cells. (A) Endogenous Cdc20 was immunoprecipitated from the lysates of either exponentially growing HeLa cells (Interphase) or from HeLa cells obtained by mitotic ‘shake-off’ following treatment with nocodazole (3 μM for 18 h) (Mitotic). The control contained lysis buffer only. The Cdc20 immunoprecipitants were subsequently western blotted with either a Cdc20 antibody (upper panel), Mad2A antibody (middle panel), or a Mad2B antibody (lower panel). (B) Exponentially growing HeLa cells were either transfected with control siRNA, or Mad2B siRNA (5-100 nM). Cells were lysed 36 h after transfection and the cell lysates were immunoblotted with an antibody to either Mad2B (upper panel), Mad2A (middle panel), or γ-tubulin (lower panel). (C) Flow cytometry DNA profiles of HeLa cells transfected with either control siRNA (5 nM for 36 h) or Mad2B siRNA (5 nM for 48 h) and subsequently treated with nocodazole (3 μg/mL for 18 h). Molecular weight markers (kDa) are indicated on the left of each western blot (WB).

Mad2B does not inhibit APC/C activity in vitro

To test whether the binding of Mad2B to Cdc20 inhibits APC/C activity, we generated GST-Mad2B and GST-Mad2A, cleaved the GST-tag, and purified recombinant proteins to near homogeneity, as shown in Fig. 3A. First, purified Mad2B and Mad2A proteins were tested for their ability to bind Cdc20 in vitro. Cdc20 protein was immunoprecipitated from HeLa cell lysates, and the immunoprecipitants were incubated with the purified Mad2B and Mad2A proteins over a range of concentrations (0.4-8.0 nM). Subsequent western blotting of the Cdc20 immunoprecipitants indicated that Cdc20 had a higher affinity for Mad2A at low concentrations of Mad2A (0.4-0.8 nM); however, at a higher concentration (8 nM) both Mad2A and Mad2B displayed a similar affinity for Cdc20 (Fig. 3B). Cdc20 did not bind to the purified GST protein over the tested concentration range. To analyze the functional impact of Mad2B binding to Cdc20, we performed an in vitro ubiquitylation assay to assess APC/C activity. As shown in Fig. 3C, the activation of APC/C in this assay was Cdc20-dependent (compare lanes 1 and 2). Recombinant Mad2A, Mad2B, and GST alone did not affect the APC/C activity (lanes 3-5). In the presence of full-length BubR1, Mad2A (1.5 µM) strongly inhibited APC/C activity by 78% in comparison with 16.3% inhibition by Mad2B (1.5 µM) or 1.5 µM GST (30.6% inhibition). A previous study has demonstrated that a fragment of BubR1 (amino acids 1-370) is sufficient for APC/C inhibition in vitro (Lara-Gonzalez et al., 2011). When we performed the in vitro ubiquitylation assay in the presence of BubR1 (1-370), Mad2A (1.5 µM) inhibited APC/C activity by 92% compared to the 7% inhibition by Mad2B (1.5 µM) and 4.4% inhibition by GST (1.5 µM). This result suggested that although purified Mad2B is capable of binding Cdc20 as effectively as Mad2A, this interaction does not inhibit APC/C activity, at least in vitro.

Fig. 3.

Recombinant Mad2B does not inhibit the APC/C in vitro. (A) Coomassie blue-stained gel of GST-Mad2A, Mad2B, and GST proteins following purification as described in the Materials and Methods. (B) Purified, recombinant Mad2A, Mad2B, and GST proteins (0.4-8 nM) were incubated with immunoprecipitated Cdc20 for 1 h at 4°C. The Cdc20 immunoprecipitants were subsequently western blotted with antibodies to Mad2A, Mad2B, GST, and Cdc20. For control IP’s the Cdc20 antibody was incubated with lysis buffer only. The HeLa cell extract was loaded as a positive control. (C) Western blot of an in vitro ubiquitylation assay probed with an anti-Myc tag antibody to detect cyclin B1 N90 conjugates (upper panel). The same blot was then re-probed with an anti-Cdc27 antibody. The level of cyclin B1 N90 ubiquitylation was normalized to the amount of Cdc27 protein and the quantitated data is shown in the lower panel (each bar is the mean of two independent experiments). Molecular weight markers (kDa) are indicated on the left of each western blot (WB).

Mad2B colocalizes with PCNA and γ-H2AX at sites of cisplatin-induced DNA damage

As Mad2B/Rev7 is a subunit of DNA Pol ζ, which is known to play a role in the mammalian DDR, we treated HeLa cells with cisplatin to activate the DNA damage checkpoint (Liu et al., 2003). Cisplatin treatment caused a time-dependent increase in the phosphorylation of endogenous Chk1 (a direct target of ATM and ATR kinases) (Guo et al., 2000; Jazayeri et al., 2006), as assessed by western blotting with a phospho-Chk1 (Ser317) antibody (Fig. 4A). Cisplatin treatment has also been reported to cause activation of PCNA, phosphorylation of H2AX (Ser139) (γ-H2AX), and accumulation of both proteins at foci of DNA damage (Hicks et al., 2010). We examined whether Mad2B is also recruited to DNA damage sites following cisplatin treatment in HeLa cells overexpressing HA-Mad2B. In untreated control cells, HA-Mad2B was present in the nucleus; however, it did not display any immunoreactivity with either PCNA or γ-H2AX antibodies (Fig. 4B, 4C, upper panels). Following cisplatin treatment, redistribution of HA-Mad2B in the nucleus occurred, which was found to co-localize with PCNA (Fig. 4B, lower panel) and γ-H2AX (Fig. 4C, lower panel) at DNA damage foci. The accumulation of Mad2B in the PCNA and γ-H2AX-foci involved the redistribution of the existing pool of Mad2B and the synthesis of the Mad2B protein. Western blot analysis of cytoplasmic and chromatin-associated proteins indicated that Mad2B was primarily a chromatin-associated protein (Fig. 4D). Furthermore, nuclease treatment of HeLa cells overexpressing HA-Mad2B resulted in the complete loss of HA-Mad2B from the nucleus (Fig. 4E), confirming the chromatin association of Mad2B. To examine changes in Mad2B protein expression following cisplatin treatment, immunoblot analysis was performed on total HeLa cell lysates at intervals following cisplatin addition. The results of western blotting indicated that there was a gradual increase in the level of Mad2B protein following cisplatin addition (Fig. 4F).

Fig. 4.

Mad2B is a chromatin-associated protein that colocalizes with the DNA damage markers PCNA and γ-H2AX. (A) Immunoblot analysis of the phosphorylation of Chk1, using a phospho-Chk1 (Ser 317) antibody, following treatment of exponentially growing HeLa cells with cisplatin (50 μM) for the indicated times (top panel). Parallel cell lysates were immunoblotted with either a total Chk1 antibody (middle panel) or with a γ-tubulin antibody (lower panel). (B, C) HA-Mad2B colocalizes with PCNA and γ-H2AX following cisplatin addition. HeLa cells were transfected with HA-Mad2B for 48 h and subsequently treated with cisplatin for 2 h. Cells were first extracted and then fixed and stained for HA-Mad2B (green), DNA (blue), and either PCNA (red) or γ-H2AX (red). Scale bar, 10 μm. (D) Mad2B binding to chromatin increased following cisplatin treatment. Exponentially growing HeLa cells were treated with cisplatin for 12 h and the cytoplasmic and chromatin-associated protein fractions immunoblotted with antibodies to Mad2B, Histone H3, and α-tubulin. Scale bar, 10 μm. (E) HeLa cells were transfected with HA-Mad2B for 48 h and subsequently treated with micrococcal nuclease (0.008 U/μL) for 30 min, extracted, and fixed and stained for HA-Mad2B (green), DNA (blue), and α-tubulin (red). (F) Cisplatin (50 μM) increased the expression of Mad2B as determined by western blotting of whole cell extracts. Molecular weight markers (kDa) are indicated on the left of each western blot (WB).

siRNA-mediated depletion of Mad2B enhances cisplatin-induced chromosomal damage and apoptosis

As Mad2B (hRev7) is involved in the DDR following DNA damage, we next examined the effect of the loss of Mad2B on morphological changes in the nuclei. Immunofluorescence analysis of HeLa cells following treatment of Mad2B-depleted cells with cisplatin indicated a significant increase in the frequency of cells displaying chromatin during damage (Fig. 5). Nuclear defects included cells with toroidal nuclei, micronuclei, degenerated nuclei, and fragmented nuclei (Fig. 5B), and quantification of these phenotypes (Fig. 5C) indicated that there was a significant increase in the frequency of damaged nuclei in Mad2B-depleted cells (30%) compared to the control oligonucleotide-treated cells. In addition, cisplatin treatment increased the number of damaged nuclei in HeLa cells (Untreated), and Mad2B depletion displayed an increase in damaged nuclei compared to control siRNA transfection in the absence of cisplatin treatment (Fig. 5C).

Fig. 5.

Inactivation of the Mad2B protein results in various forms of nuclear breakage after cisplatin-induced DNA damage. HeLa cells were either mock transfected (Untreated), transfected with the control (CTL) siRNA oligonucleotide (10 nM), or transfected with Mad2B siRNA oligonucleotides (10 nM). After transfection for 48 h, cells were treated with cisplatin (50 μM) for a further 36 h. DNA was stained with Hoechst 33342 (blue). (A) Control-siRNA treated cell. (B) Mad2B siRNA-treated cell with damaged nuclei. Arrows show micronuclei, degenerated nuclei, and fragmented apoptotic nuclei. Scale bar indicates 20 μm. (C) Quantitation of the data shown in (A, B). Approximately 200 cells were counted in randomly selected fields and the number of cells displaying damaged nuclei is shown. Each bar represents the mean ± SD of three independent experiments. *Indicates p<0.05, **indicates p<0.01 using a Student’s t-test.

To test whether Mad2B knockdown induces apoptosis by cisplatin-induced DNA damage, cells were stained with an M30 antibody using immunofluorescence analysis. M30 measures caspase-cleaved cytokeratin 18 (CK18) cytoskeletal protein during apoptosis (Roth et al., 2004). Previous studies have reported that the M30 antibody detects only apoptotic cells, and not viable and necrotic cells, by immunohistochemistry, and that reactivity of the M30 antibody is associated with the apoptosis index (Roth et al., 2004; Schutte et al., 2004). Immunofluorescence analysis indicated a significant increase in M30 staining apoptosis in HeLa cells following the treatment of Mad2B-depleted cells with cisplatin (Fig. 6). In addition, cisplatin treatment increased the number of apoptotic cells (CTL oligo), and Mad2B depletion displayed an increase in M30 staining apoptotic cells compared to control cells (Untreated) in the absence of cisplatin treatment (Fig. 6B). These results indicated that the loss of Mad2B may accelerate DNA damage in human cancer cells upon cisplatin treatment, which displays a chromatin damage phenotype, thereby leading to apoptosis.

Fig. 6.

Loss of the Mad2B protein causes apoptosis following cisplatin-induced DNA damage. HeLa cells were either mock transfected (Untreated), transfected with the control (CTL) siRNA oligonucleotide (10 nM), or transfected with Mad2B siRNA oligonucleotides (10 nM). After transfection for 48 h, cells were treated with cisplatin (50 μM) for a further 36 h, stained with cytokeratin 18 cleavage (M30) antibody (green), and DNA was stained with Hoechst 33342 (blue) (A). Scale bar indicates 10 μm. (B) Quantitation of the data shown in (A). Approximately 100 cells were counted in randomly selected fields and the number of cells displaying damaged nuclei is shown. Each bar represents the mean ± SD of three independent experiments. *p<0.05, **p<0.01 using a Student’s t-test.

DISCUSSION

The data obtained in the present study led us to conclude that Mad2B is not a mitotic checkpoint protein in mammalian cells for several reasons. First, in eukaryotic cells, mitotic checkpoint proteins such as Mad1, Mad2A, Bub1, Bub3, BubR1(Mad3), and Mps1 localize to the kinetochores during mitosis, where they monitor the attachment of kinetochores to spindle microtubules (Musacchio and Salmon, 2007; Sacristan and Kops, 2015). Mad2B is not recruited to kinetochores upon SAC activation, indicating that Mad2B, in contrast to Mad2A, localizes to the cytoplasm in mitotically arrested HeLa cells (Fig. 1). This is consistent with the finding that Mad2B does not bind to Mad1 (Chen and Fang, 2001). Binding of Mad2A to Mad1 is essential for the recruitment of Mad2A to unattached kinetochores (Sironi et al., 2002). Second, one mechanism by which SAC activation suppresses anaphase onset is through the kinetochore-dependent activation of Mad2A (Nezi et al., 2006). Our biochemical analyses showed that Mad2B does not bind to Cdc20 upon nocodazole activation of the SAC (Fig. 2A). Additionally, we confirmed that Mad2B levels were uniform throughout the cell cycle (Supplementary Fig. 1), and Cdc20 had a lower affinity for Mad2B at low concentrations of Mad2B than Mad2A in the interphase cell lysate (Fig. 3B). Third, siRNA-mediated depletion of Mad2B did not impair SAC function (Fig. 2C). Finally, the results of the in vitro ubiquitylation assay indicated that Mad2B, in contrast to Mad2A, did not inhibit the activity of mitotic APC/C, indicating that cyclin B was hardly ubiquitylated (Fig. 3C). Mad2B has also been reported to have no effect on the activity of the interphase APC/C (Fang, 2002).

The function of Mad2B (Rev7 in yeast) has been studied in both yeast and mammalian cells, and it has been proposed that Rev3 and Rev1 regulate mutagenic translesional DNA synthesis (Gibbs et al., 2005; Okada et al., 2005). The data presented in this study are consistent with the role of Mad2B in DNA damage and repair.

Treatment with cisplatin led to the activation of Chk1, resulting in phosphorylation at the Ser 317 residue (Fig. 4A). Additionally, it was confirmed that γ-H2AX foci, a DNA damage marker, were formed (Fig. 4C), and the number of apoptotic cells increased in cisplatin-treated cells (Fig. 5, 6). Current models (Lehmann, 2011; Sharma et al., 2013) describing the mechanism of cisplatin-induced DNA damage indicate the recruitment of numerous proteins, including endonucleases and DNA repair polymerases, to the sites of DNA damage. Thus, we confirmed that cisplatin treatment in cells triggers a DDR.

We found that Rev7/Mad2B is a chromatin-associated protein (Fig. 4D, 4E) that colocalizes with markers of DNA damage, such as PCNA and γ-H2AX (Fig. 4B, 4C, respectively). Additionally, Mad2B expression increased in HeLa cells after cisplatin-induced DNA damage (Fig. 4D, 4F). Thus, Rev7/Mad2B may participate in DNA damage repair at the site of DNA damage, in response to cisplatin treatment. Other studies have also demonstrated the co-localization of Rev1 with either PCNA or γ-H2AX following cisplatin treatment (Hicks et al., 2010). Central to lesion bypass by Pol ζ (a complex of Rev3 and Rev7) and strong binding to Rev1 is the recruitment of the E3 ubiquitin ligase Rad6 (E2)-Rad18 (E3) to the site of DNA damage where it ubiquitylates PCNA (Leung et al., 2019). In addition, it was confirmed that Mad2B/Rev7 foci were formed during DNA damage induced by cisplatin, recruited to γ-H2AX foci, and Mad2B/Rev7 was co-localized with γ-H2AX (Fig. 4C). γ-H2AX functions to amplify the factor for DDR signaling at DNA damage sites to repair DNA, thus, it is an indicator of DNA DSBs (Chen et al., 2019; Collins et al., 2020). Thus, our results suggested that Mad2B/Rev7 functions in the DNA damage repair pathway in response to cisplatin-induced DNA damage.

To demonstrate the importance of Mad2B in DNA repair, we silenced the expression of Mad2B using siRNA. Although we did not observe any changes in cell cycle progression in the absence of Mad2B, treatment of Mad2B-depleted HeLa cells with cisplatin resulted in a significant increase in the frequency of cells with chromatin fragmentation (Fig. 5). This is consistent with a previous study that reported that shRNA-mediated depletion of Mad2B in nasopharyngeal carcinoma cells in response to DNA damage increases the frequency of chromosomal breaks (Cheung et al., 2006). There were various forms of nuclear DNA damage following the depletion of Mad2B in cisplatin-treated HeLa cells (Fig. 5). Cisplatin-induced chromosomal damage could be a form of apoptosis in human cells (Kashyap et al., 2015). Thereafter, we showed that depletion of Mad2B induced apoptosis in HeLa cells via cisplatin-induced DNA damage (Fig. 6). Specific proteolytic cleavage of CK18 is an event that occurs before the disruption of membrane asymmetry, and DNA strand breaks occur (Roth et al., 2004). This probably explains why the proportion of apoptotic cells stained with M30 (Fig. 6) was lower than the proportion of damaged nuclei (Fig. 5). These results indicated that the loss of Mad2B may accelerate DNA damage in human cancer cells upon cisplatin treatment, which displays a chromatin damage phenotype, thereby leading to apoptosis. Moreover, the results suggest that Mad2B depletion causes dysfunction of DNA damage repair machinery so it contributes to the development of genetic instability in human cancer in response to cisplatin-induced DNA damage. Thus, this result reveal that inhibition of Mad2B increases genotoxic anticancer drug cisplatin sensitivity in human cervical cancer HeLa cell. However, further studies are needed to determine whether Mad2B binds to Rev3 and Rev1 via binding to PCNA in the DNA damage response.

In summary, Mad2B, unlike Mad2A, does not function in the SAC because it did not localize to the kinetochores during mitosis, did not bind to Cdc20 upon activation of the SAC, and did not inhibit the activity of mitotic APC/C in nocodazole-treated HeLa cells. However, Mad2B functions in cisplatin-induced DDR, demonstrating that Mad2B co-localizes with PCNA and γ-H2AX, and that Mad2B/Rev7 depletion causes accumulation of damaged nuclei, thereby accelerating apoptosis in human cancer cells in response to cisplatin-induced DNA damage. Therefore, Mad2B may be a critical modulator of DDR.

Funding Statement

ACKNOWLEDGMENTS We would like to thank Dr. Pablo Lara-Gonzalez to help in vitro ubiquitylation assay. This work was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) (NRF-2022R1A2C1012144 to J.H.K.).