Old Passengers as New Drivers: Chromosomal Passenger Proteins Engage in Translesion Synthesis

Abstract

Survivin is known for its dual biological role in apoptosis inhibition and mitotic progression. In addition to its being part of the chromosomal passenger complex (CPC), recent findings suggest additional roles for Survivin in the DNA damage response, further contributing to therapy resistance. In this study, we investigated the role of Survivin and the CPC proteins in the cellular response to irradiation with a focus on DNA replication processes. As is known, ionizing radiation leads to an increased expression of Survivin and its accumulation in nuclear foci, which we now know to be specifically localized to centromeric heterochromatin. The depletion of Survivin and Aurora B increases the DNA damage marker γH2AX, indicative of an impaired repair capacity. The presence of Survivin and the CPC in nuclear foci that we already identified during the S phase co-localize with the proliferating cell nuclear antigen (PCNA), further implying a potential role during replication. Indeed, Survivin knockdown reduced replication fork speed as assessed via DNA fiber assays. Mechanistically, we identified a PIP-box motif in INCENP mediating the interaction with PCNA to assist in managing damage-induced replication stress. Survivin depletion forces cells to undergo unphysiological genome replication via mitotic DNA synthesis (MiDAS), resulting in chromosome breaks. Finally, we revealed that Aurora B kinase liberates Pol η by phosphorylating polymerase delta-interacting protein 2 (POLDIP2) to resume the replication of damaged sites via translesion synthesis. In this study, we assigned a direct function to the CPC in the transition from stalled replication forks to translesion synthesis, further emphasizing the ubiquitous overexpression of Survivin particularly in tumors. This study, for the first time, assigns a direct function to the chromosomal passenger complex, CPC, including Survivin, Aurora B kinase, Borealin, and INCENP, in the transition from stalled replication forks (involving PCNA binding) to translesion synthesis (liberating Pol η by phosphorylating POLDIP2), and thus in maintaining genomic integrity.

Keywords: Aurora B kinase; DNA damage response; Survivin; replication; translesion synthesis.

Figure 1

Irradiation increases Survivin expression and relocation to nuclear foci as part of the CPC. (A,B) A431 cells stably expressing Survivin-GFP (A), as well as parental A431 cells (B), were irradiated with 6 Gy, fixed 30 min, and 24 h after IR, permeabilized and subjected to confocal microscopy. Cells in (B) were additionally immunostained with a Survivin-specific antibody (green). DNA was stained with Hoechst (blue). Non-irradiated cells (w/o IR) were used as controls. Scale bar: 7.5 μm. (C) A431 cells stably expressing Survivin-GFP were irradiated with 1 or 6 Gy and lysed 1 h after IR for immunoblot analysis. (D) A431 cells stably expressing Survivin-GFP were irradiated with 6 Gy and, 1 h after IR, subjected to subcellular fractionation, separating cytoplasm (CP) from nuclei (NU), and subsequent immunoblot analysis. GAPDH served as loading control for the CP fraction and Lamin A/C for the NU fraction, respectively. The relative signal intensities of Survivin bands were quantified via a densitometric analysis using ImageJ and normalized to GAPDH (C) or GAPDH and Lamin A/C (D). The depicted values represent the relative increase in Survivin levels after IR. (E–G) A431 cells stably expressing Survivin-GFP were irradiated with 6 Gy, fixed 1 h after IR, and permeabilized. Cells were immunostained with antibodies specific to Aurora B (E), Borealin (F), or INCENP (G), respectively (red), and subjected to confocal microscopy. DNA was stained with Hoechst (blue). Non-irradiated cells (w/o IR) were used as controls. Scale bar: 7.5 μm. The insets show higher magnifications of the areas outlined in the main panels. Representative images are shown.

Figure 2

Survivin accumulates in centromeric regions after irradiation. (A) Schematic representation of centromeric heterochromatin. In contrast to transcriptionally silenced heterochromatin, mainly characterized by Heterochromatin protein-1 (HP1) bound to the methylated lysine 9 residue of histone H3 (H3K9me), constitutive or centromeric heterochromatin additionally relies on CENP-A as a stable centromere component and CENP-B that also belongs to the centromere and is also recognized by CREST. (B) A431 and HeLa cells were co- transfected with plasmids coding for EGPF-HP1α (green) and Survivin-tdTomato (red), permeabilized, fixed, and subjected to confocal microscopy. DNA was stained with Hoechst (blue). Scale bar: 7.5 μm. (C,D) A431 cells stably expressing Survivin-GFP (C), as well as parental A431 cells (D), were irradiated with 6 Gy, fixed 30 min and 24 h after IR, permeabilized, and immunostained with CREST serum, detecting centromeric chromatin (red) for confocal microscopy. Cells in (D) were additionally incubated with a Survivin-specific antibody (green). DNA was stained with Hoechst (blue). Non-irradiated cells (w/o IR) were used as controls. Scale bar: 7.5 μm. The insets show higher magnifications of the areas outlined in the main panels. Representative images are shown.

Figure 3

H2AX phosphorylation increases after Survivin depletion. (A,B) A431 cells were transfected with 20 nM of non-targeting control siRNA (siCtl) and Survivin-specific siRNA (siSurv), irradiated with 6 Gy 48 h after transfection, and fixed 24 h after IR. As indicated in the scheme, the phosphorylation of histone H2AX occurs in the vicinity of damaged DNA. Cells were immunostained with antibodies specific to phosphorylated H2AX (γH2AX, pS139) (red) and Survivin (green) and analyzed via confocal microscopy in two different magnifications. DNA was stained with Hoechst (blue). Non-irradiated cells (w/o IR) were used as a control. The insets show higher magnifications of the areas outlined in the main panels. Representative images are shown. Scale bars: 25 μm (A) and 7.5 μm (B). (C) For quantification, images were processed with ImageJ. Each bar represents the mean ± SD of residual γH2AX foci per nucleus of two repeated experiments (200–300 evaluated nuclei each) of cells treated with non-targeting control siRNA (siCtl) or Survivin-specific siRNA (siSurv).

Figure 4

Interaction of PCNA with the CPC enables unhindered replication fork progression. (A) WI-38 cells were fixed, permeabilized, and immunostained for Survivin (green) and PCNA (red). DNA was stained with Hoechst (blue). Scale bar: 5 µm. Insets, higher magnifications of the areas outlined in the main panels. Left: schematic representation of characteristic patterns of PCNA distribution in S-phase substages. (B,C) Survivin knockdown reduces replication fork speed. A431 cells were transfected with siRNA specific to Survivin (siSurv), non-targeting siRNA (siCtrl), or without siRNA (mock) as a control. Then, 72 h after transfection, cells were sequentially pulse-labeled with CldU and IdU for 20 min each. Cells were harvested, and the DNA fiber assay was performed. The length of stained fibers from elongating forks was measured. Results are depicted as a frequency distribution. Means ± SD from three independent experiments with 300 fibers analyzed per experiment are shown. (C) Dot plot of the mean replication fork speed of one representative data set out of three independent experiments, as shown in (B). Bars: means ± SD (n = 300 fibers of each group). ****: p < 0.0001. (D,E) CPC proteins interact with PCNA as analyzed via a proximity ligation assay (PLA, (D)) in HeLa cells (E). Representative images are shown. Each red spot represents a single PLA interaction signal. DNA was stained with Hoechst (blue). Scale bar: 5 µm. The number of PLA signals per nucleus from >150 cells was quantified with ImageJ. Bars: means ± SEM of a positive control (dark blue), negative controls (light blue), and interaction samples (yellow). ****: p < 0.0001. (F–H). INCENP binds to PCNA via a PIP-box motif. Exemplary cartoon presentation of a PCNA monomer (blue) as part of the homotrimer (transparent) bound to the p21 PIP-box peptide (brown, 1AXC) (F). Overview of the PIP-box consensus, the same motif in wildtype INCENP (wt) and the mutated (PIPmut) PIP-box motif (mutated aa, highlighted in bold) (G). 293T cells were transfected (or co-transfected) with GFP-PCNA, myc-INCENPwt, or myc-INCENP PIPmut, as indicated (H). Single transfections are highlighted in dark blue. Lysates were immunoprecipitated (IP) with magnetic α-GFP beads (yellow) or uncoupled beads (light blue) and immunoblotted with indicated antibodies. Representative images are shown.

Figure 5

Survivin depletion impedes the cell’s response to replication stress. (A). Subcellular expression of the CPC after induction of replication stress; 293T cells were treated with different reagents inducing replication stress, including 1 µM of Camptothecin, 0.5 mM of Hydroxyurea, and IR with 10 Gy. Whole-cell extracts: and subcellular fractions were generated and subjected to immunoblot analysis with indicated antibodies. γH2AX served as a marker for DNA damage and repair, and Tubulin and Histone H3 served as loading controls for the cytoplasmic fraction and the chromatin fraction, respectively. (B). CPC depletion results in mitotic DNA synthesis. HeLa cells were transfected with siRNAi against Survivin or Aurora B, fixed, and permeabilized 48 h later. During the last 16 h of depletion, cells were incubated with RO-3306 to arrest them in G₂ phase. Cells were stained with the Click-iT™ EdU Alexa Fluor™ 488 Imaging kit. DNA was stained with Hoechst (blue). Scale bar: 25 µm. Quantitative analysis of EdU foci in metaphase cells. Bars: mean number of EdU foci per cell ± SD in >50 cells. Data were analyzed via a t-test. *: p < 0.05. n = 3. (C). Survivin depletion leads to an increased number of breaks in metaphase chromosomes (black arrows). Brightfield images of metaphase spreads with a full set of chromosomes. Scale bar: 10 µm. Quantitative analysis of breaks/gaps per metaphase and of the number of metaphases with breaks. Bars: mean number of breaks per metaphase or the mean number of metaphases with breaks ± SD in >100 metaphases. Data were analyzed via a t-test. **: p < 0.01. n = 3. (D,E) Survivin depletion leads to an increase in 53BP1 nuclear bodies during the G₁ phase. HeLa cells were transfected with Survivin-specific or non-targeting siRNA, fixed, and permeabilized after 48 h. Cells were immunostained with antibodies specific to 53BP1 (green) and cyclin A (red). DNA was stained with Hoechst (blue). Scale bars: 50 and 10 µm (D). Quantitative analysis of 53BP1 nuclear bodies in cells treated with Survivin-specific or non-targeting siRNA (E). Bars: mean number of 53BP1 nuclear bodies per nucleus ± SD in >100 cells. Data were analyzed via a t-test. **: p < 0.01; *: p < 0.05. n = 3.

Figure 6

POLDIP2 is a substrate of Aurora B kinase, and it links the CPC to translesion synthesis. (A) Expression of polymerase η after depletion of CPC members and concurrent irradiation. HeLa cells were transfected with non-targeting, Survivin-specific, or Aurora B-specific siRNA and irradiated with 10 Gy 48 h later. Whole-cell extracts were generated and subjected to immunoblot analysis with indicated antibodies. Tubulin served as a loading control. (B,C) Co-immunoprecipitation reveals an interaction between CPC members and polymerase η. 293T cells were co-transfected with plasmids coding for FLAG-Pol η, together with the indicated myc-tagged CPC members. Then, 24 h later, cells were irradiated with 10 Gy and incubated for 2 h. Prior to immunoprecipitation, chromatin extraction was performed. FLAG-Pol η was immunoprecipitated from the extracts, and analysis was performed with the indicated antibodies. (D) POLDIP2 is a substrate of Aurora B kinase. Aurora B kinase assay performed with the known substrate MBP and POLDIP2. ADP production mirrors enzymatic activity. Both data sets have been fit with sigmoidal curves (POLDIP2: R2 = 0.99; MBP: R2 = 0.98). (E) Three-dimensional model of POLDIP2 (Q9Y2S7) from the AF2 database. ATR phosphorylation sites (S147, S150) are labeled, as well as possible Aurora B phosphorylation sites (T5, T216) according to PhosphoPICK. (F) Detail image of residues in close proximity of T216 (left). Several hydrophobic residues are located in the surrounding of T216 facing directly towards this residue. Following the phosphorylation of T216, distances to hydrophobic amino acids decrease (right). (G) Co-immunoprecipitation reveals a decreased interaction between POLDIP2 T216E and Pol η. 293T cells were transfected with plasmids encoding POLDIP2 wildtype-, T216E-, or AuroraB-GFP. Lysates were immunoprecipitated (IP) with Pol η-specific antibodies and immunoblotted for GFP. (H) Model of the CPC’s role in translesion synthesis. Upon DNA damage, PCNA is ubiquitinated at stalled replication forks and the CPC binds to PCNA via INCENP’s PIP-box motif. Sequestered Pol η is liberated after phosphorylation of POLDIP2 via Aurora B kinase. Due to the high affinity and close proximity, Pol η displaces Pol δ and resumes replication via translesion synthesis.