ATM controls DNA repair and mitochondria transfer between neighboring cells
- 40 Downloads
Intercellular communication is essential for multicellular tissue vitality and homeostasis. We show that healthy cells message protective signals through direct cell–cell connections to adjacent DNA–damaged cells in a microtubule–dependent manner. In DNA–damaged cells, mitochondria restoration is facilitated by fusion with undamaged mitochondria from healthy cells and their DNA damage repair is optimized in presence of healthy cells. Both, mitochondria transfer and intercellular signaling for an enhanced DNA damage response are critically regulated by the activity of the DNA repair protein ataxia telangiectasia mutated (ATM). These healthy–to–damaged prosurvival processes sustain normal tissue integrity and may be exploitable for overcoming resistance to therapy in diseases such as cancer.
KeywordsCell–cell communication Genotoxic stress Mitochondria exchange DNA damage repair Ataxia telangiectasia mutated (ATM)
DNA double–strand breaks
Non–homologous end joining
Reactive oxygen species
Super–resolution radial fluctuation
In multicellular organisms, neighboring cells continuously exchange information for coordinating homeostasis, survival and development. Direct communication through cell–cell contact is one of the most succinct ways to exchange information  as demonstrated by transfer of small organelles like mitochondria carrying signaling molecules such as RNA [2, 3, 4]. Under stress conditions, mitochondria often move from one cell to another and between different cell types [1, 5, 6, 7]. Limited mitochondrial functionality results from exposure to DNA–damaging agents like ionizing radiation, which is rescued by mitochondria fission and fusion [8, 9].
A prominent process for intercellular communication is known as bystander effect  describing a DNA damage response in healthy cells triggered by signals from damaged cells . Among the various types of DNA damage, DNA double–strand breaks (DSB), as most life–threatening lesions, are repaired through non–homologous end joining (NHEJ) and homologous recombination (HR). Upon DSB sensing , ataxia–telangiectasia mutated (ATM) and DNA protein kinase catalytic subunit (DNA–PKcs) phosphorylate numerous substrates including histone H2AX and p53 binding protein 1 (53BP1) [13, 14, 15]. A variety of signaling molecules such as ATM , nitric oxide (NO) and reactive oxygen species (ROS) have been shown to be relayed through gap junctions between DNA damaged and healthy cells . Remaining is the important and yet open question how healthy cells influence DNA–damaged cells.
Here, we designed a co–culture model system containing human pancreatic cancer cells and ATMwt as well as ATM−/− fibroblasts, in which genotoxically injured cells were co–cultured with healthy cells upon removal of secreted factors putatively connectable to bystander effects. We comparatively characterized the DSB repair in target and non–target cell populations under co– and mono–culture conditions. Our results show that the presence of healthy cells profoundly modifies the DNA damage response of genotoxically injured cells by a microtubule– and ATM–dependent exchange of healthy mitochondria.
Materials and methods
Human pancreatic cancer cells (MiaPaCa–2), wild–type (wt) and ATM−/− fibroblasts, MiaPaCa–2–GFP, MiaPaCa–2–G1G2, ATMwt–GFP and ATM−/−–GFP were grown in Dulbecco’s Modified Eagle Medium (DMEM, without phenol red, Gibco) containing 10% (v/v) fetal bovine serum, 1% (v/v) non–essential amino acids, 4 mM L–Glutamine, 100 units/ml penicillin–streptomycin, 4.5 g/L D–glucose. Cells were maintained at 37 °C with 5% CO2. All cells were tested negative for mycoplasma. MiaPaCa–2 cells were purchased from the American Type Culture Collection (ATCC). Human fibroblasts (ATMwt; ATM−/−) were kindly provided by P. A. Jeggo (University of Sussex, UK) .
Plasmid DNA transfection
Cells were plated onto uncoated 35 mm dishes with a 0.17 mm glass bottom (MatTek) and allowed to reach 60–70% confluency. GFP empty vector or Fucci cell cycle indicator series (Amalgaam) were introduced into the cells using Lipofectamine 2000 (Invitrogen,) according to the manufacture’s protocol. Briefly, cells were incubated in 250 μl OptiMEM with a DNA–mix of 250 μl (containing plasmid DNA of 1–2 μg / μl and 10 μl Lipofectamine 2000 in 240 μl OptiMEM). Transfection media was removed after 4 h and cells were further incubated in fresh medium. The expression of plasmid DNA was examined by an epi–fluorescence microscope. To generate stably transfected cells, cells were selected using culture media containing Geneticin (Gibco) at 300–800 μg / ml or Hygromycin B (Invitrogen) at 300 μg / ml over 2 weeks.
Treatment with pharmacological inhibitor
Taxol (Paclitaxel, Sigma, stock 10 mM in ethanol), Colchicine (Sigma, stock 50 mg/ml in ethanol), ATM inhibitor (Calbiochem, 118,500, stock 10 mM in DMSO)  and DNA–PK inhibitor (NU7026, Selleckchem, 324,788, stock 10 mM in DMSO)  were applied to cells 45 min prior to radiation exposure at 10 μM, 50 nM and at indicated concentrations, respectively.
DNA damage induction using x–ray radiation
Cells were trypsinized and transferred either into a falcon tube as suspension culture or onto 35 mm glass bottom culture dishes coated with Poly–D–Lysine (MatTek). Irradiation was delivered while cells were in suspension at room temperature using 6 Gy single doses of 200 kV x–rays (Yxlon Y.TU 320; dose rate∼1.3 Gy/min at 20 mA) filtered with 0.5 mm Cu as published . The absorbed dose was measured using a Duplex dosimeter (PTW).
Total protein extractions and Western blotting
One day after plating cells onto 6–well plates or at 1 h after radiation exposure, cells were harvested by scraping using cell lysis buffer (Cell Signaling). Total protein concentration (mg/ml) of cell lysate was determined using a spectrophotometry (Nanodrop, Thermo Fisher). Samples were used immediately or stored at − 80 °C until use. Proteins were separated by sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE) prior to Western blotting. Detection of specific proteins was performed with enhanced chemiluminescence reagent (Amersham) after antibody binding using LAS–3000 imaging system (Fuji).
Co–culture system and fluorescent staining in living cells
Immunofluorescence (IF) staining was conducted as published . Briefly, cells were fixed with 3.7% paraformaldehyde at room temperature (RT) for 10 min and permeabilized with 0.5% Triton X–100 at RT for 5 min. Then, cells were incubated with blocking buffer (3% BSA in PBS) at RT for 1 h, followed by incubation with primary and secondary antibody in blocking buffer in the dark at RT for 1 h, respectively. The cells were stored in the dark at 4 °C until being imaged.
Epifluorescence and spinning disc confocal microscopy
All the microscopic experiments were performed in 35 mm culture dishes with glass bottom. For image–based analysis with large cell numbers as shown in Additional file 1: Figure S1, samples were screened by Zeiss Axio Observer Inverted microscope using a 10x/ objective. For confocal microscopy, the images were recorded by spinning disc confocal microscopy (Olympus IX83 microscope with Yokogawa CSU–X1 Confocal Spinning Disc unit, and iXon Ultra 897 EMCCD Camera) with Olympus Plan S Apo 40x/0.95 or 60x/1.2 NA water immersion objective. Confocal fluorescence images were obtained by sequential excitation, if more than two fluorophores were monitored. The fluorescence was measured for DAPI and Hoechst 33342 with excitation wavelength of 405 nm and emission wavelengths 425/45 nm; Alexa 488, MitoTracker Green FM and GFP with excitation wavelength of 488 nm and emission wavelengths 525/20 nm; Alexa 546 and 594, CellTracker Red CMTPX and MitoTracker Red CMXRos with excitation wavelengths of 561 nm and emission wavelengths 617/73 nm; Alexa 633 and 647, DRAQ5 and MitoTracker Deep Red FM with excitation wavelengths of 640 nm and emission 685/40 nm. Phase–contrast images were recorded simultaneously in indicated experiments. Z–stacks with a step of 1 μm were acquired per image.
Super–resolution radial fluctuation analysis
Images for super–resolution radial fluctuation (SRRF) analysis were acquired by spinning disc confocal microscopy as described above. To yield a super–resolution frame, a stack of 200 raw data images were processed by running SRRF program with a ring radius of 1. The SRRF algorithm plugin for Fiji is provided by Gustafsson and colleagues .
Primary antibodies: anti–53BP1 (IF 1:400, WB 1:1000, Novus Biologicals, NB100–904), anti–ATM (WB 1:1000, Cell signaling, 2873), anti–beta actin (WB 1:10,000, Abcam, ab8224), anti–phospho–ATM S1981 (WB 1:500, Rockland, 200–301–400), anti–phospho–DNA–PK S2056 (IF 1:300, Abcam, ab18192), anti–phospho–Histone H2A.X S193 (γH2AX, IF 1:200, anti–rabbit, Cell signaling, 9718 and 1: 300, anti–mouse, Millipore, 05–636), anti–Tubulin (IF 1:500, Abcam, ab6160). Secondary antibodies for IF were purchased from Life Technologies (working dilution at 1:500 in blocking buffer). Those antibodies were validated in the presence or without protein of interest. HRP–conjugated secondary antibodies for WB were from GE Healthcare (working dilution at 1:5000).
All results represent mean ± standard deviation (SD), at least three independent experiments were performed. Unpaired, two–sided Student’s t–test was performed by Microsoft Excel. A p–value is less than 0.5 was considered significant.
Results and discussion
Bilateral transfer of mitochondria between acceptor and donor cells
To investigate intercellular communication between DNA–damaged and non–damaged cells, we deployed a co–culture system enabling us to explore how healthy cells serve as donors for prosurvival cues for DNA–damaged acceptor cells (Fig. 1a). To test for efficient repair and restoration of damaged mitochondria by functional mitochondria from healthy cells, we tracked mitochondria originating from unirradiated (donor) and irradiated (acceptor) fibroblasts (Fig. 1b and Additional file 1: Figure S1). We discovered an intensive bilateral exchange accompanied by high level of colocalization (Fig. 1c). Disturbing microtubules known to serve as mitochondria carriers by the microtubule–stabilizing agent taxol  conserved bilateral mitochondria exchange in contrast to the microtubule polymerization blocker colchicine  (Additional file 2: Figure S2). Analyzing mitochondria shapes revealed induced formation of tubular structures (high aspect ratio (AR) value) and fusion between donor and acceptor mitochondria in co–culture (Fig. 1d–f and Additional file 3: Figure S3, Additional file 4: Figure S4). Hence, irradiation–induced fragmentation causative for mitochondria dysfunctionality seemed to be repaired by mitochondria exchange and fusion from healthy donor cells. This process depended on an existing microtubule network as taxol but not colchicine enabled more tubular, less fragmented mitochondria.
Based on intercellular trafficking of DNA repair proteins like ataxia telangiectasia mutated (ATM) along microtubules  and ATM involvement in mitochondria homeostasis and redox–sensing [29, 30], we next determined the relevance of ATM for mitochondria exchange upon DNA damage. In contrast to co–cultures of ATM wildtype (wt) fibroblasts (Additional file 5: Figure S5a and b), co–cultures of unirradiated ATM−/− with irradiated ATM−/− fibroblasts lacked mitochondria transfer (Additional file 5: Figure S5f).
ATM–deficient cells are incapable of mitochondria delivery to neighboring cells suggesting mitochondria transfer to follow a one–way, here ATMwt–to–ATM−/− fibroblasts, course of action independent from whether ATM−/− fibroblasts act as acceptor or donor (Additional file 5: Figure S5). The strong spreading ability of MitoT Red has often been observed [31, 32]. However, in our hands, MitoT Red originating from ATM−/− cells could not be found in neighboring cells (Additional file 5: Figure S5 g). We, therefore, carried out further control experiments by using different combinations of MitoT (Additional file 5: Figure S5 g), and, different ratios of donor/acceptor cell numbers (Additional file 5: Figure S5 h). Moreover, the laser intensity for detecting ATM−/− cells was appropriately overexposed. Taken together, our data suggest that ATM functionality seemed to govern the ability for intercellular mitochondria transfer,as one of the foundations for later fusion, independent of the status of damage.
Accelerated DNA repair through presence of healthy cells
Co–culturing modulates cell cycle distribution
We next analyzed simultaneously cell cycling and DSB as tightly and cooperatively intertwined processes . In contrast to mono–culture, the presence of healthy cells resulted in two separable acceptor cell fractions upon DNA damage: G2–phase cells with high and G1–phase cells with low 53BP1 foci numbers (Additional file 8: Figure S8). These data demonstrate minimal differences in cell cycling of DNA damaged cells when co–cultured with healthy cells relative to mono–cultures as well as a cell cycle phase–dependent association with DSB.
Cell–to–cell signaling stimulated the DNA damage response in ATM deficient cells
ATM deficient cells are incapable of rescuing DNA damaged neighboring cells
Next, we examined whether ATM signaling is fundamental in healthy donors for rescuing damaged neighboring cells. A functional ATM status determined an appr. 50% DSB reduction in irradiated MiaPaCa–2 cells co–cultured with healthy, ATM–competent fibroblasts relative to mono–culture, co–culture with ATM–incompetent fibroblasts or pharmacologically inhibited ATM (Fig. 3c). As ATM activates both error–prone NHEJ and error–free HR repair , we employed NHEJ/HR–proficient (MiaPaCa–2, ATMwt fibroblasts), NHEJ–deficient (ATM−/− fibroblasts) and HR–deficient cells (BRCA2–depleted Capan1) as donors and found that, when co–cultured, NHEJ/HR–proficient MiaPaCa–2 acceptor cells show optimized DSB repair dependent on NHEJ and functional ATM but not HR (Fig. 3d). These observations suggest a prominent role of ATM for donor–to–acceptor but not acceptor–to–donor cell signaling to provide specific, yet to be identified cues, for a more effective NHEJ–related DSB repair.
Intercellular signaling requires microtubule polymerization
Cells communicate constantly with their environment including neighboring cells as well the extracellular matrix by giving and receiving signals. Studies over the past decades have provided ample evidence on how unirradiated cells are damaged through communication with neighboring irradiated cells. Very little attention, however, has been paid on the reverse, i.e. feedback signals from healthy cells to DNA damaged cells. The prosurvival and DNA repair modulating effects transferred from undamaged to damaged cells are largely unknown. For both normal and cancer tissue, it would be fundamental to clarify whether this feedback exists, especially when target cell signals fail to trigger a DNA damage response in non–target cells, what are such signals, how are these signals delivered to target cells and what is the response pattern of target cells.
Here, we addressed the DNA damage response in genotoxically injured cells cultured in the presence or absence of undamaged cells. Collectively, our data show that (i) bilateral transfer of mitochondria between target (irradiated) and non–target cells occurs in an ATM–dependent manner; (ii) DNA damage repair in target (irradiated) cells is accelerated through the presence of non–target cells in both ATMwt and ATM deficient fibroblasts; (iii) cell cycle distribution in target cells is substantially different in co–cultured target cells as compared to mono–cultures of target cells; and (iv) a functional microtubules system is required for this intercellular signaling through direct cell–to–cell contact.
These findings demonstrate that cell–cell communication between DNA–damaged and healthy cells provides effective and immediate prosurvival feedback. The hierarchical organization between microtubules, mitochondria and ATM pinpoints the complexity of such regulatory networks allowing multiple ways for rescue and evasion. While this is fundamental for normal tissue integrity, rescue signals may particularly be considered in various diseases such as cancer, in which inherent and acquired resistance mechanisms to therapies impede the cure of patients. Given the interactive and regulatory role of these three target structures, it seems promising to search for combinatorial therapy approaches that abrogate prosurvival feedback between damaged and undamaged cell populations.
We thank all the members of the Core Facility Cellular Imaging (CFCI) at Faculty of Medicine Carl Gustav Carus, Technische Universität Dresden for technical assistance.
SJ and NC conceived, designed the study, analyzed as well interpreted the Data, and wrote the manuscript; SJ performed the experiments; NC supervised the project. Both authors read and approved the final manuscript.
This work was supported by the German Research Foundation [FKZ: CO 668/8–19] and the Faculty of Medicine Carl Gustav Carus (MeDDrive program), Technische Universität Dresden.
Ethics approval and consent to participate
Consent for publication
The authors declare that they have no competing interests.
- 7.Vignais ML, Caicedo A, Brondello JM, Jorgensen C. Cell connections by tunneling nanotubes: Effects of mitochondrial trafficking on target cell metabolism, homeostasis, and response to therapy. Stem Cells Int. 2017:6917941.Google Scholar
- 10.Nagasawa H, Little JB. Induction of sister chromatid exchanges by extremely low doses of a-particles. Cancer Res. 1992;15:6394–6.Google Scholar
- 17.Azzam EI, de Toledo SM, Little JB. Direct evidence for the participation of gap junction-mediated intercellular communication in the transmission of damage signals from alpha-particle irradiated to nonirradiated cells. Proc Natl Acad Sci. 2012;98:473–8.Google Scholar
- 31.Cselenyák A, Pankotai E, Csordás A, Kiss L, Lacza Z. Live-Cell Fluorescent Imaging of Membrane or Mitochondrion Transfer between Connected Cells in Culture. Microsc Sci Technol Appl Educ. 2010:764–71.Google Scholar
- 34.Blackford AN, Jackson SP. ATM, ATR, and DNA-PK: The Trinity at the Heart of the DNA Damage Response. Mol. Cell. 2017:801–17.Google Scholar
- 36.Pesce CG, Zdraljevic S, Peria WJ, Bush A, Repetto MV, Rockwell D, et al. Single-cell profiling screen identifies microtubule-dependent reduction of variability in signaling. Mol Syst Biol. 2018;14.Google Scholar
Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.