Multiple Facets of the DNA Damage Response Contribute to the Radioresistance of Mouse Mesenchymal Stromal Cell Lines

Abstract

The regeneration of the hematopoietic system following total body irradiation is supported by host-derived mesenchymal stromal cells (MSCs) within the bone marrow. The mechanisms used by MSCs to survive radiation doses that are lethal to the hematopoietic system are poorly understood. The DNA damage response (DDR) represents a cohort of signaling pathways that enable cells to execute biological responses to genotoxic stress. Here, we examine the role of the DDR in mediating the resistance of MSCs to ionizing radiation (IR) treatment using two authentic clonal mouse MSC lines, MS5 and ST2, and primary bulk mouse MSCs. We show that multiple DDR mechanisms contribute to the radio-resistance of MSCs: robust DDR activation via rapid γ-H2AX formation, activation of effective S and G₂/M DNA damage checkpoints, and efficient repair of IR-induced DNA double-strand breaks. We show that MSCs are intrinsically programmed to maximize survival following IR treatment by expressing high levels of key DDR proteins including ATM, Chk2, and DNA Ligase IV; high levels of the anti-apoptotic, Bcl-2 and Bcl-_XL; and low levels of the pro-apoptotic, Bim and Puma. As a result, we demonstrate that irradiated mouse MSCs withstand IR-induced genotoxic stress, continue to proliferate, and retain their capacity to differentiate long-term along mesenchymal-derived lineages. We have shown, for the first time, that the DDR plays key roles in mediating the radioresistance of mouse MSCs which may have important implications for the study and application of MSCs in allogeneic bone marrow transplantation, graft-versus-host disease, and cancer treatment.

γ-Radiation, Mesenchymal stromal cells, DNA damage response, DNA damage checkpoints, DNA repair, Apoptosis

Introduction

The adult bone marrow contains two systems of stem cells known as hematopoietic stem cells (HSCs) and mesenchymal stem cells/mesenchymal stromal cells (MSCs). The main function of the bone marrow is to produce mature functional blood cells that circulate in the peripheral blood. Peripheral blood cells are under continuous renewal which is sustained by the production of hematopoietic progenitors by HSCs. Hematopoietic homeostasis is the physiological state in which the rates of hematopoietic cells entering and leaving the peripheral blood are balanced. The maintenance of hematopoietic homeostasis is dependent on the interaction between HSCs and the HSC niche within the bone marrow. Within the HSC niche, MSCs provide soluble factors, contact-dependent interactions, and physical support to control HSC self-renewal, quiescence, and differentiation [1, 2]. MSCs have been traditionally defined as bone marrow-derived spindle-shaped cells that rapidly adhere to plastic and proliferate in vitro to form individual colonies [3, 4]. MSCs are characterized by the presence of cell surface proteins including CD73, CD44, CD90, and CD105 and by the absence of hematopoietic cell surface markers including CD45 and CD14 [4]. MSCs are also multipotent stem cells that can differentiate along various lineages to become specialized cell types including osteocytes, adipocytes, and chondrocytes [3].

When the hematopoietic system is exposed to ionizing radiation (IR), for example, γ-rays and x-rays, hematopoietic homeostasis is disturbed. Hematopoietic stem and progenitor cells are extremely radiosensitive (D₀ of 0.6–1.6 Gy) [5]. Lymphopenia rapidly occurs following acute exposure to even low doses of radiation, whereas mature granulocytes, erythrocytes, and platelets can tolerate higher doses [5]. However, following myeloablative radiation exposure, the depletion of hematopoietic stem and progenitor cells causes the pools of differentiating and maturing blood cells to become exhausted, leading to hematopoietic failure [5, 6]. Bone marrow transplantation (BMT) is currently the only treatment available for hematopoietic failure following total body irradiation. The destruction of the host hematopoietic system by a lethal radiation dose creates a compartment within the bone marrow in which donor HSCs and their progenitors can engraft and give rise to a new hematopoietic system. Several studies have shown that MSCs support the reconstitution of the hematopoietic system following BMT [7–10]. In addition, it has been demonstrated that while patient-derived hematopoietic cells following allogeneic BMT are of donor origin, MSCs isolated from these patients are of host origin [11, 12]. These findings indicate that MSCs are able to survive doses of radiation that are lethal to the hematopoietic system.

The major trigger of the cellular response to IR is its destructive impact on genome integrity. Cells can mount a co-ordinated response to genotoxic stresses, including IR, by activating a network of interacting signaling pathways, collectively known as the DNA damage response (DDR). The DDR signaling pathways consist of (a) sensor proteins that recognize sites of damaged DNA and activate signal cascades by interacting with (b) transducer proteins that relay the signal to other downstream transducer proteins and, in turn, to (c) effector proteins that act on various genes and other proteins to induce a biological response [13–15]. These effector proteins are involved in mechanisms that can cause cells to undergo cell cycle arrest, initiate DNA repair, become senescent, or, in the presence of irreparable DNA damage, to undergo apoptosis [15]. Therefore, the ability of cells to effectively execute DDR signaling is essential for restoring genomic stability and for promoting survival following DNA damage.

The mechanisms underlying the radio-Resistance of MSCs are currently poorly understood. A major element hindering advancement in this area is the use of bulk MSC populations that are heterogenous in nature and vary in phenotype between isolations. Therefore, we have compared primary bulk mouse MSCs with two cloned mouse MSC lines, MS5 and ST2, which are widely used as supportive stroma for HSCs and their progenitors in vitro [16, 17]. In recent years, they have been confirmed to be authentic MSC lines that express MSC cell surface markers and differentiate along mesenchymal-derived lineages [18–22]. The objective of our study was to examine the role of the DDR in mediating the resistance of MSCs to gamma radiation (γ-radiation). We have compared in detail the DDR of the mouse MSC lines, MS5 and ST2, with that of the radio-sensitive mouse CD4⁺ CD8⁺ thymocyte cell line, ST4.5, to determine key aspects of the DDR that contribute to the radioresistance of MSCs. This is the first study where the DDR of cloned mouse MSCs has been investigated in detail.

Materials and Methods

Cell Culture and Treatment

Multiple independent isolates of C57BL/6 primary bulk mouse MSCs were provided by Prof. Matthew D. Griffin (REMEDI, NUI Galway) and cultured from passages 5 to 8, as previously described [23]. MS5 and ST2 MSC lines were provided by Prof. Antonius Rolink (Department of Biomedicine, University of Basel) and were re-cloned prior to use. Primary bulk MSCs and MSC lines were CD45⁻, CD44⁺, and CD29⁺ (supporting information Fig. S1). ST4.5 CD4⁺ CD8⁺ thymocyte cell line was provided by Dr. Anne Wilson (Ludwig Institute of Cancer Research, Lausanne) and J774A.1 monocyte/macrophage cell line by Prof. Benjamin Bradley (Department of Orthopaedics, University of Bristol). All cell lines were cultured in Dulbecco's modified Eagle's medium high glucose (Sigma-Aldrich, St. Louis, http://www.sigmaaldrich.com) supplemented with 10% fetal bovine serum (FBS) (Lonza, Walkersville, MD, http://www.lonza.com) and 1% penicillin/streptomycin sulfate solution. All cell types were cultured in a humidified incubator at 37°C containing 5% CO₂. Cells were γ-irradiated at the indicated doses using a Mainance Millennium Sample Irradiator containing a ¹³⁷Cs source at a dose rate of approximately 102 cGy/minute. Cells were treated with 1 μM Staurosporine solution (Cell Signaling Technologies, Beverly, MA, http://www.cellsignal.com) and harvested at the indicated time points post-treatment.

Clonogenic Survival Assay

Adherent cells were irradiated at the indicated doses and seeded into six-well plates (Nunc, Rochester, NY, http://www.nuncbrand.com) at a concentration of 200–500 cells per well, depending on cell type. Cells were incubated for 7–14 days until colonies of each cell type were clearly visible. Colonies were stained with Coomassie Blue (Sigma-Aldrich) and counted. Non-adherent ST4.5 cells were seeded into T25 flasks (Nunc) at a concentration of 5,000 cells per milliliter, harvested 5 days post irradiation, and cell numbers were counted in duplicate using a hemocytometer. The percentage survival of each cell type was determined by normalizing the number of colonies per cell generated by irradiated cultures to the number of colonies per cell generated by control cultures. The plating efficiency of each cell type was MS5 (56.4% ± 10%), ST2 (60.8% ± 12.5%), J774A.1 (47% ± 4.85%), and bulk MSCs (27.9% ± 1.1%).

Differentiation Assays

Control and irradiated (10 Gy) cells were seeded into 24-well plates (Nunc) at a concentration of 30,000 cells per well and of 5,000 cells per well for performance of osteogenic and adipogenic assays, respectively. Osteogenic and adipogenic differentiation capacities were determined using standard assays [24]. All images were captured using an Olympus IX71 Inverted Fluorescent Microscope with Olympus Cell∧P Software (Olympus, Hamburg, Germany, http://www.olympus-global.com).

Antibodies

For Western blotting, anti-phospho-Histone H2A.X (Ser139) mouse monoclonal antibody (Millipore, Billerica, MA, http://www.millipore.com), anti-H2AX rabbit polyclonal antibody, anti-ATM [2C1(1A1)] mouse monoclonal antibody, anti-DNA Ligase IV rabbit polyclonal antibody, anti-Ku70 [N3H10] mouse monoclonal antibody, anti-Ku80 [5C5] mouse monoclonal antibody, anti-β-Tubulin rabbit polyclonal antibody (Abcam, Cambridge, U.K., http://www.abcam.com), anti-ATR (N-19) goat polyclonal antibody, anti-Chk2 (H-300) rabbit polyclonal antibody (Santa Cruz Biotechnology Inc., Santa Cruz, CA, http://www.scbt.com), anti-p53 1C12 mouse monoclonal antibody, anti-poly(ADP-ribose) polymerase (PARP) rabbit polyclonal antibody, anti-Caspase-3 rabbit polyclonal antibody, anti-Bcl-_XL (54H6) rabbit monoclonal antibody, anti-Bcl-2 (D17C4) rabbit monoclonal antibody, anti-Bim (C34C5) rabbit monoclonal antibody, anti-Puma rabbit polyclonal antibody (Cell Signaling Technologies), anti-β-Actin rabbit polyclonal antibody (Sigma-Aldrich), anti-DNA-PK_cs Ab-4 mouse monoclonal antibody, Pierce horseradish peroxidase (HRP)-conjugated rabbit anti-mouse IgG antibody, and ImmunoPure HRP-conjugated goat anti-rabbit IgG antibody (Thermo Scientific, Rockford, IL, http://www.thermoscientific.com) were used. For immunofluorescence staining, anti-phospho-Histone H2A.X (Ser139) mouse monoclonal antibody (Millipore) and anti-Rad51 rabbit polyclonal antibody (Abcam) were used. Secondary fluorescein (FITC)-conjugated AffiniPure F(ab′)₂ Fragment goat anti-mouse IgG antibody and Texas Red-conjugated AffiniPure F(ab′)₂ Fragment goat anti-rabbit IgG antibody (Jackson ImmunoResearch Laboratories Inc., West Grove, PA, http://www.jacksonimmuno.com) were used. For flow cytometry, anti-5′-bromodeoxyuridine (BrdU) mouse monoclonal antibody (BD Biosciences, San Diego, http://www.bdbiosciences.com) and FITC-conjugated anti-mouse IgG (whole molecule) antibody (Sigma-Aldrich) were used. Annexin-V-FITC conjugate was a gift from Prof. Corrado Santocanale (Centre for Chromosome Biology, NCBES, NUI Galway).

Western Blotting

Whole cell extracts were prepared from harvested control or irradiated cells at the indicated time points post irradiation by resuspending cell pellets in 4× SDS-loading buffer. Samples were heated and sonicated prior to separation using SDS-PAGE gels and transferred to nitrocellulose membranes. Chemiluminescence was detected using SuperSignal West Pico Chemiluminescent Substrate (Thermo Scientific) and medical x-ray film (Konica Minolta Medical & Graphic Imaging Inc., Ramsey, NJ, http://www.konicaminolta.com).

Immunofluorescence Staining

MSCs were cultured on glass coverslips (Fisher Scientific, Hampton, NH, http://www.fisherscientific.com) prior to irradiation. Irradiated ST4.5 cells were transferred to poly(L-lysine)-coated microscope slides (Fisher Scientific) using a cytospin. All cultures were fixed in 4% paraformaldehyde (Sigma-Aldrich), permeabilized in 0.1% Triton X-100 solution (Sigma-Aldrich), and incubated with the indicated primary antibodies at 4°C overnight. Following incubation with the indicated secondary antibodies at 37°C for 1 hour, cells were mounted in Vectashield containing Hoechst solution (Sigma-Aldrich). All images were captured using an IX71 Olympus fluorescent microscope with ImageProPlus 6.0 software (Media Cybernetics, Crofton, MD, http://www.mediacy.com). Manual quantification of γ-H2AX foci per cell (total of 50 cells per time point) was performed blind and was further confirmed by automated quantification using Cell Profiler software [25].

Flow Cytometry

Cell Cycle Analysis Using BrdU–Propidium Iodide Staining

Cells were pulsed for 15 minutes–4 hours (depending on cell type) with 25 μM BrdU (Sigma-Aldrich), washed in PBS, and resuspended in growth medium. Cells were harvested at the indicated time points post-irradiation and fixed in ice-cold 70% ethanol. Following acid denaturation using HCl, the cells were blocked in phosphate buffered saline (PBS) containing 0.1% Triton X-100 solution and 0.5% bovine serum albumin (Sigma-Aldrich). The cells were sequentially incubated with anti-BrdU antibody (BD Biosciences) and FITC-conjugated anti-mouse IgG antibody (Sigma-Aldrich) for 45 minutes each at room temperature, separated by a washing step, and then resuspended in propidium iodide/RNase staining buffer (BD Biosciences). The progression of G₁, S, and G₂/M cells through the cell cycle was analyzed by measuring the percentage of cells in each phase until 72 hours post irradiation.

Analysis of Apoptosis Using Annexin-V–Propidium Iodide Staining

Cells were irradiated and harvested at the indicated time points post-irradiation. Adherent cells were incubated at 37°C for 5 minutes. All cell types were then resuspended in Annexin-V binding buffer (10 mM HEPES, 150 mM NaCl, 5 mM KCl, 1 mM MgCl₂, and 1.8 mM CaCl₂, pH 7.4) and incubated in FITC-conjugated Annexin-V solution for 15 minutes on ice. Five microliters of 20 μg/l propidium iodide (PI) solution (Sigma-Aldrich) was added to each cell suspension prior to analysis. All fluorescence-activated cell sorting (FACS) samples were analyzed using a BD FACS Canto Flow Cytometer and FlowJo software (TreeStar Inc., OR, http://www.flowjo.com).

In Vitro DNA Double-Strand Break End-Joining Assay

Briefly, nuclear extracts were prepared according to manufacturer's instructions (Active Motif, La Hulpe, Belgium, http://www.activemotif.com). The pmaxFP-Green-N (pMAX-Green Fluorescent Protein) plasmid (Lonza) was linearized by cleavage with Xmn1. End-joining efficiency of nuclear extracts was measured as previously described [26].

Results

MSCs Are Resistant to IR Treatment

To compare the radiosensitivity of MSCs with hematopoietic cells, MSC lines (MS5 and ST2), primary bulk MSCs, and hematopoietic cell lines (J774A.1 and ST4.5) were irradiated at 2–10 Gy and cultured at low density for a period of 5–14 days (depending on cell type). At the highest IR dose of 10 Gy, ∼21% MS5 cells, ∼34% ST2 cells, and ∼12% primary bulk MSCs formed colonies in contrast with 0.65% J774A.1 and 0% ST4.5 cells (Fig. 1). No viable ST4.5 cells were detected above 6 Gy irradiation (* in Fig. 1). The LD₅₀ dose was 2 Gy for ST4.5, 4 Gy for J774A.1, and 6 Gy for MSCs (Fig. 1). Overall, these results indicate that MSCs are substantially more radioresistant than hematopoietic cells. They also confirm that the MSC lines, MS5 and ST2, exhibit levels of radioresistance similar to that of primary bulk MSCs. The HSC niche where MSCs reside in vivo is a relatively hypoxic environment with O₂ levels ranging from 2% to 8% [27]. Interestingly, we found that growth in hypoxia (5% O₂) (a) induced an approximately two-fold increase in MSC survival at 6 and 8 Gy, (b) increased the LD₅₀ of MSC lines to 8 Gy, (c) enhanced MSC growth, and (d) increased colony density (supporting information Fig. S2). These results confirm that MSC radio-resistance is maintained under hypoxic conditions.

Figure 1.

Mouse MSCs are resistant to γ-radiation treatment. Clonogenic survival assay of mouse MSC lines (MS5 and ST2), primary bulk mouse MSCs, and J774A.1 and ST4.5 cell lines γ-irradiated at 2–10 Gy and cultured for 5–14 days (depending on cell type) as previously described. Error bars represent mean ± SD, n = 3. *, Viable ST4.5 cells were not detected above 6 Gy irradiation. Abbreviations: IR, ionizing radiation; MSC, mesenchymal stromal cell.

MSCs Rapidly Activate the IR-Induced DDR

Confirming previous studies, we have shown that ST4.5 cells are sensitive to IR treatment [28, 29]. Therefore, using ST4.5 as a model for radiosensitivity, we examined the contribution of DDR mechanisms to the radioresistance of MSCs. The ability of MSCs and ST4.5 cells to activate the DDR was determined by analyzing γ-H2AX formation at 0–24 hours post irradiation (post IR) using Western blotting. Maximal γ-H2AX expression was detected in MS5 and ST2 cells within 1 hour post 1 and 10 Gy irradiation, whereas it was at 2 hours in ST4.5 cells (Fig. 2A). We next examined whether the difference in γ-H2AX induction kinetics between irradiated MSC lines and ST4.5 cells was correlated with varying expression levels of several key DDR proteins. No difference in the expression levels of H2AX and ATR was detected between cell types (Fig. 2C; supporting information Fig. S3A). However, ATM, DNA-PK_cs, and Chk2 were expressed at higher levels in MSCs than in ST4.5 cells (Fig. 2B, 2C).

Figure 2.

Mouse MSCs rapidly activate the DNA damage response. (A): Western blot analysis of γ-H2AX expression in MS5, ST2, and ST4.5 cells at 0–24 hours post 1 and 10 Gy irradiation. (B): Western blot analysis of ATM and DNA-PK_cs expression and (C) of H2AX and Chk2 expression in control MS5, ST2, bulk MSCs, and ST4.5 cells. β-Actin and β-Tubulin expression were used as loading controls. All Western blot images are representative of one of two to three independent experiments. Abbreviations: IR, ionizing radiation; MSC, mesenchymal stromal cell.

MSCs Undergo Cell Cycle Recovery Following Activation of IR-Induced DNA Damage Checkpoints

We examined the ability of MS5 and ST4.5 cells to activate IR-induced DNA damage checkpoints using a flow cytometry-based BrdU incorporation assay. Since BrdU is only incorporated into replicating DNA (i.e., S phase cells) at the time of BrdU treatment, BrdU-PI staining enabled us to clearly identify cells in G₁ (BrdU negative), S (BrdU positive), and G₂/M (BrdU negative) phases (Fig. 3A, upper left panel). We observed dramatic differences in the response of the BrdU-positive populations of MS5 and ST4.5 following 10 Gy treatment. The irradiated BrdU positive MS5 population exhibited delayed cell cycle progression, indicative of the activation of intra-S-phase and G₂ checkpoint mechanisms, as the proportion of BrdU-positive cells remained essentially unchanged until 12 hours post IR in comparison with untreated cells (Fig. 3B). In contrast, the BrdU positive ST4.5 population reduced from ∼47.9% to ∼32.1% within 12 hours post IR compared with ∼53% to ∼23.7% in the control (Fig. 3B). A major factor contributing to this reduction in the ST4.5 cell line was the generation of sub-G₁ cells beginning at 2 hours post IR which originated from both the BrdU negative and positive populations (Fig. 3A, arrowheads in lower panels). Irradiated BrdU labeled MS5 and ST4.5 cells that progressed through S phase subsequently accumulated as a cohort in G₂/M until 12 hours post IR (Fig. 3A, right panels, and Fig. 3B) and exhibited delayed re-entry into G₁, indicating activation of G₂/M accumulation checkpoint (supporting information Fig. S3B) [30]. Following low-dose irradiation (1 Gy), transient delays in S and G₂/M phase progression followed by a 2-hour delay in G₁ re-entry were observed in both MS5 and ST4.5 cells, indicating that the duration of DNA damage checkpoint activation was dose-dependent in both cell types (supporting information Fig. S3C, S3D).

Figure 3.

Mouse MSCs continue to cycle following DNA damage checkpoint activation. (A): Representative cytograms of MS5 and ST4.5 cells harvested at 0–72 hours post 10 Gy irradiation and stained for BrdU incorporation and DNA content using PI. Black boxes indicate G₁ and G₂/M (BrdU negative) and S (BrdU positive) populations. Black arrowheads indicate sub-G₁ cells originating from BrdU negative and positive populations. *, indicates cohort of BrdU labeled cells accumulated at the G₂/M checkpoint and ⋄ indicates BrdU labeled G₁ cells. (B): Graph of the percentage BrdU labeled S phase cells of MS5 and ST4.5 at 0–72 hours post 10 Gy irradiation. Error bars represent mean ± SD, n = 2 (control) or n = 3 (irradiated). (C): Western blot analysis of p53 stabilization in MS5, ST2, and ST4.5 cells harvested at 0–4 hours post IR at 1 and 10 Gy using β-Tubulin expression as a loading control. All flow cytometry and Western blot images are representative of one of three independent experiments. Abbreviations: BrdU, 5′-bromodeoxyuridine; IR, ionizing radiation; PI, propidium iodide.

The BrdU labeled MS5 population remained arrested at the G₂/M checkpoint until 12 hours post IR (Fig. 3A, upper right panel), whereas the BrdU labeled ST4.5 population was maintained at this checkpoint until 24 hours post IR (Fig. 3B; supporting information Fig. S3B). Approximately 42% of total BrdU labeled MS5 cells was present in G₁/S at 24 hours post IR, whereas BrdU labeled ST4.5 cells entered the sub-G₁ population (supporting information Fig. S3A, S3B). In addition, BrdU labeled G₁ MS5 cells re-initiated DNA synthesis, indicated by a reduction in BrdU signal intensity at 72 hours post IR (Fig. 3A, upper right panel). ST2 cells could not be analyzed by flow cytometry due to the high autofluorescence of this cell type. However, irradiated (10 Gy) BrdU labeled primary bulk MSCs also progressed as a cohort through S phase, accumulated at the G₂/M checkpoint and exhibited delayed re-entry into G₁, indicating activation of S and G₂/M DNA damage checkpoints (supporting information Fig. S4A, S4B). In addition, approximately 48% of total BrdU labeled MSCs was present in G₁/S at 36 hours post IR with only low levels of sub-G₁ cells detected.

To examine the G₁/S checkpoint response of these cell types, we analyzed p53 stabilization at 0–4 hours post IR. We found that p53 was only transiently stabilized by irradiated MS5 and ST2 cells, whereas p53 stabilization was maintained in irradiated ST4.5 cells (Fig. 3C). After 10 Gy irradiation, a dose-dependent increase in p53 stabilization was detected in MS5 and ST2 cells, whereas p53 stabilization kinetics remained unaltered in ST4.5 cells (Fig. 3C). Irradiated primary bulk MSCs displayed transient p53 stabilization kinetics similar to the MSC lines and stabilization was also more pronounced following 10 Gy irradiation (supporting information Fig. S4C).

MSCs Are Resistant to IR-Induced Apoptosis

The rapid appearance of sub-G₁ cells and prolonged p53 stabilization by irradiated ST4.5 cells, in contrast to MSCs, indicated that these cell types may execute different apoptotic responses following IR treatment. Annexin-V/PI staining was performed to confirm whether irradiation induced apoptosis in MSCs and ST4.5 cells. At only 4 hours post IR, ∼84.6% ST4.5 cells were Annexin-V positive compared with ∼1.7% MS5 cells (Fig. 4A, 4B). At 36 hours post IR, only ∼4.7% MS5 cells were Annexin-V positive in comparison with ∼96% ST4.5 cells (Fig. 4B).

Figure 4.

Mouse MSCs are resistant to IR-induced apoptosis. (A): Representative cytograms of MS5 and ST4.5 cells harvested at 0–36 hours post 10 Gy irradiation and stained with Annexin-V and for DNA content using PI. (B): Graph of the percentage Annexin-V positive MS5 and ST4.5 cells at 0–36 hours post 10 Gy irradiation. Error bars represent mean ± SD, n = 3. (C): Western blot analysis of Caspase-3 and PARP cleavage in MS5, ST2, and ST4.5 cells harvested at 0–36 hours post IR (10 Gy) and Staurosporine treatment (1 μM). (D): Western blot analysis of Bcl-2 and Bcl-_XL and (E) of Bim isoforms and Puma expression in control MS5, ST2, bulk MSCs, and ST4.5 cells. β-Tubulin expression was used as a loading control. All flow cytometry and Western blot images are representative of one of three independent experiments. Abbreviations: IR, ionizing radiation; MSC, mesenchymal stromal cell; PARP, poly(ADP-ribose) polymerase; PI, propidium iodide; Puma, p53 upregulated modulator of apoptosis.

To determine whether MSCs activate the intrinsic apoptotic pathway in response to IR-induced DNA damage, we analyzed the cleavage of pro-Caspase-3 and PARP post 10 Gy irradiation. Minimal levels of Caspase-3 and cleaved PARP were detected in irradiated MS5 and ST2 cells (Fig. 4C, left panels). In contrast, Caspase-3 was detected in irradiated ST4.5 cells beginning at 4 hours post IR with levels remaining high for the duration of the experiment (Fig. 4C, left panels). PARP cleavage was detected in untreated ST4.5 cells, which is likely due to the normal presence of a proportion of apoptotic cells correlated with the rapid turnover of this cell type. Nevertheless, a striking increase in PARP cleavage was detected in ST4.5 cells following irradiation (Fig. 4C, left panels). To confirm that MSCs contain a functional intrinsic apoptotic pathway, all three cell lines were treated with 1 μM Staurosporine that potently induced pro-Caspase-3 and PARP cleavage in all cell types (Fig. 4C, right panels).

To further examine the apparent resistance of MSCs to IR-induced apoptosis, we analyzed the basal expression levels of pro-apoptotic and anti-apoptotic proteins in MSCs and ST4.5 cells. We found that MSCs express higher levels of Bcl-2 and Bcl-_XL (Bcl-extra long) than ST4.5 cells (Fig. 4D). In addition, MSCs expressed very low levels of Bim (Bcl-2 interacting protein) and Puma (p53 upregulated modulator of apoptosis) in comparison with ST4.5 cells (Fig. 4E).

MSCs Resolve IR-Induced DNA Double-Strand Breaks

The ability of irradiated MSCs to dephosphorylate H2AX and to resume proliferation without significant apoptosis suggested that MSCs may have a high capacity for repairing IR-induced DNA damage. To examine whether irradiated MSCs can potentially repair DNA double-strand breaks (DSBs), we analyzed the resolution of γ-H2AX IR-induced foci (IRIF) by MSCs and ST4.5 cells following 1 and 10 Gy irradiation (Fig. 5A, 5B). Greater numbers of γ-H2AX IRIF were detected in MS5 and ST2 cells than in ST4.5 cells at 1 hour post IR (Fig. 5A, 5B; supporting information Fig. S5A, S5B). This corresponded with the faster induction of γ-H2AX formation detected in irradiated MSC lines relative to ST4.5 cells (Fig. 1A). γ-H2AX IRIF were largely resolved in MS5 and ST2 cells at 24 hours post 1 and 10 Gy irradiation, which also correlated with the reduction in γ-H2AX signal detected in the MSC lines at this time point (Fig. 1A). In contrast, γ-H2AX IRIF persisted in irradiated ST4.5 cells (Fig. 5A–5C; supporting information Fig. S5A).

Figure 5.

Mouse MSCs repair DNA double-strand breaks. (A, B): Graph of the average number of γ-H2AX foci per cell (50 cells in total per time point) at 0, 1, and 24 hours post 1 and 10 Gy irradiation. Error bars represent mean ± SD, n = 2. (C): Representative immunofluorescent images of MS5, ST2, and ST4.5 cells at 0–24 hours post 10 Gy irradiation and stained for DNA content (DAPI), γ-H2AX (green), and Rad51 (red) IRIF captured using ×60 magnification. (D): Western blot analysis of Rad51 expression in control MS5, ST2, bulk MSCs, and ST4.5 cells. (E): Analysis of efficiency in end-joining reactions containing linearized plasmid and nuclear extracts/nuclear extracts denatured at 95°C for 5 minutes/cytoplasmic extract derived from control MS5, ST2, and ST4.5 cells. (F): Western blot analysis of DNA ligase IV, DNA-PK_cs, Ku70, and Ku80 expression in nuclear extracts of control MS5, ST2, and ST4.5 cells. (G): Western blot analysis of DNA ligase IV, Ku70, and Ku80 expression in whole cell extracts of control MS5, ST2, bulk MSCs, and ST4.5 cells. β-Tubulin expression was used as a loading control. Agarose gel, immunofluorescent, and Western blot images are representative of one of three independent experiments. Abbreviations: DAPI, 4′,6-diamidino-2-phenylindole dihydrochloride; IRIF, ionizing radiation-induced foci; MSC, mesenchymal stromal cell.

In contrast to the MSC lines, highly variable numbers of γ-H2AX IRIF were detected in irradiated bulk MSCs at 1 hour post high dose (10 Gy) irradiation (supporting information Fig. S6A, S6B). In addition, conflicting data were found between the kinetics of γ-H2AX IRIF formation and γ-H2AX expression by irradiated bulk MSCs. Following 1 Gy irradiation, low γ-H2AX signal was detected at 1 hour post IR by Western blotting, whereas increased numbers of γ-H2AX foci were observed at this time point. In addition, following 10 Gy irradiation, strong γ-H2AX signal was still present at 24 hours post IR, whereas γ-H2AX foci numbers were reduced at this time point (supporting information Fig. S6A–S6C). Thus, bulk MSCs display a heterogenous response to IR-induced DNA DSBs. Importantly, bulk MSC cultures contained no CD45⁺ cells (supporting information Fig. S1). Rad51 IRIF, which colocalized with γ-H2AX IRIF, were also detected in all cell types following IR treatment and persisted in MSCs at 24 hours post 10 Gy irradiation, which may be indicative of DNA DSB repair by homologous recombination (HR) (Fig. 5C; supporting information Figs. S5A, S6A). Interestingly, MSCs were also found to express higher levels of Rad51 than ST4.5 cells (Fig. 5D).

To investigate the ability of MSCs to repair DNA DSBs using end-joining reactions, we used an established in vitro assay to examine the capacity of control nuclear extracts to religate digested plasmid DNA containing a single DNA DSB [26]. Nuclear extracts from MS5 and ST2 cells converted monomeric DNA fragments into multimers more efficiently than nuclear extracts from ST4.5 cells (Fig. 5E). Successful end-joining was inhibited by heat denaturation of nuclear extracts and was not detected in the presence of cytoplasmic extracts (Fig. 5E). DNA Ligase IV and β-Tubulin expression were enriched in nuclear and cytoplasmic extracts, respectively, demonstrating nuclear extract purity (supporting information Fig. S5C). Higher expression levels of DNA Ligase IV and DNA-PK_cs were detected in both nuclear and whole cell extracts of MS5 and ST2 compared with ST4.5 cells, which may suggest that MSCs can perform end-joining reactions using nonhomologous end-joining (NHEJ) (Fig. 5F, 5G). Taken together, these results suggest that MSCs may have a higher capacity to repair IR-induced DNA double-strand breaks than ST4.5 cells, which is likely to contribute to their greater ability to survive IR treatment.

MSCs Retain Their Differentiation Potential Following IR Treatment

Finally, we determined whether IR treatment affects the adipogenic and osteogenic differentiation potential of MSCs. Adipogenic differentiation, characterized by the formation of large spherical-shaped cells containing lipid droplets stained with oil red O, was detected in control and irradiated MS5 and ST2 cultures (supporting information Fig. S7A). Similar staining for calcium deposition using Alizarin Red S was found between control and irradiated cultures for each MSC line, indicative of osteogenic differentiation (supporting information Fig. S7B). Overall, these results indicate that MSCs retain their ability to differentiate following IR treatment.

Discussion

IR causes high levels of genotoxic stress due to the production of reactive oxygen species and of a wide range of genomic lesions including DNA DSBs. The execution of an effective DDR is critical for promoting cell survival following IR exposure. Previous studies examining the role of the DDR in driving the radioresistance of MSCs have been performed on primary, uncloned, bulk MSC cultures [31–33]. These bulk MSC cultures are heterogenous in nature, making detailed analysis difficult. Similarly, we found that primary bulk mouse MSCs display a heterogenous response to IR-induced DNA damage (supporting information Fig. S6). Our use of two widely used clonal authentic mouse MSC lines, MS5 and ST2, has enabled us to perform a detailed study of the IR response of MSCs in vitro. This is the first study where the DDR of cloned mouse MSCs has been investigated in detail. We would propose that only by studying cloned MSC lines can the DDR mechanisms of MSCs at the molecular level be understood. The MS5 cell line was originally established from irradiated long-term bone marrow cultures of C3H/HeNSlc mice, whereas the ST2 cell line was established from nonirradiated BALB/c mouse fetal liver [17, 34]. Despite their differing sources, we found that both MSC lines execute highly comparable responses to IR treatment. Furthermore, these two MSC lines are as radioresistant as primary bulk mouse MSCs (Fig. 1). In addition, their ability to differentiate is also maintained following high-dose irradiation (supporting information Fig. S7).

DNA DSBs are the most genotoxic lesions caused by IR. The activation of the DDR in response to DNA DSBs is characterized by ATM-dependent formation of γ-H2AX [35, 36]. Analysis of γ-H2AX expression demonstrated that irradiated MSC lines activate the DDR more rapidly than the radiosensitive CD4⁺ CD8⁺ thymocyte cell line, ST4.5 (Fig. 2A). In addition, MSCs express higher levels of ATM, DNA-PK_cs, and Chk2 that are key sensor and transducer proteins involved in orchestrating the DDR signaling pathways in response to IR-induced DNA DSBs (Fig. 2B, 2C) [37, 38]. Therefore, it is likely that high levels of ATM and DNA-PK_cs in MSCs promote efficient DDR activation via rapid γ-H2AX formation which is quickly amplified downstream due to the presence of high levels of Chk2, enabling MSCs to rapidly activate mechanisms, such as DNA repair and DNA damage checkpoints, enhancing survival following IR exposure (supporting information Fig. S8).

DNA damage caused by IR can activate DNA damage checkpoints that transiently delay or arrest cell cycle progression to facilitate DNA repair before entering the next phase of the cell cycle. By monitoring the progression of BrdU labeled S phase cells of MSCs, we found that MSCs activate robust intra-S-phase and G₂/M checkpoint mechanisms in response to high-dose irradiation (Fig. 3A, 3B). In addition, irradiated (10 Gy) BrdU-labeled MS5 cells undergo G₂/M checkpoint release faster than ST4.5 cells (Fig. 3B; supporting information Fig. S3B). The ability of irradiated MS5 cells to undergo G₂/M checkpoint release more rapidly than ST4.5 cells and to resume proliferation suggested that MSCs have a high capacity for repairing IR-induced DNA damage. In addition, we found that p53 is transiently stabilized in irradiated MSCs (Fig. 3C). A recent study demonstrated that transient p53 stabilization by irradiated mouse hair follicle bulge stem cells (BSCs) correlated with their high NHEJ capacity [26]. Using an in vitro DNA DSB end-joining assay applied in this study, we demonstrated that nuclear extracts from MSCs are more proficient at end-joining than extracts from ST4.5 cells (Fig. 5E) [26]. Analysis of γ-H2AX and Rad51 IRIF formation and resolution also suggested that MSCs can repair DNA DSBs using HR (Fig. 5A–5C). Moreover, MSCs express higher levels of the DNA DSB repair proteins—DNA ligase IV, DNA-PK_cs, and Rad51 than ST4.5 cells (Fig. 5D–5G). The high DNA DSB repair capacity of MSCs is likely to enhance their ability to overcome DNA damage checkpoint activation and to survive γ-radiation treatment (supporting information Fig. S8).

Previous studies have shown that the balance between the levels of proapoptotic and antiapoptotic proteins plays an important role in mediating the radioresistance of other stem cells such as HSCs and BSCs [26, 39]. Correspondingly, we found that MSCs express high levels of the anti-apoptotic proteins, Bcl-2 and Bcl-_XL, and very low levels of the potent pro-apoptotic proteins, Bim and Puma (Fig. 4D, 4E). In addition, despite containing functional apoptotic machinery, irradiated MSCs do not immediately activate the intrinsic apoptotic pathway (Fig. 4A–4C). The bias of proapoptotic and antiapoptotic proteins toward those that favor survival is likely to enhance the ability of MSCs to activate pro-survival mechanisms following DNA damage.

Clinically, the radio-resistance of MSCs has multiple implications. Total body irradiation is frequently used prior to allogeneic BMT. Depletion of MSC subtypes within mouse bone marrow impedes HSC maintenance and function showing that host MSCs are essential for promoting HSC engraftment and hematopoietic reconstitution [40, 41]. Co-transplantation of MSCs with HSCs enhances HSC engraftment [7–9]. Therefore, understanding MSC radiobiology could impact development of preconditioning regimens. MSCs are also important components of the tumor stroma to which they have been recruited [2]. Tumor cells can secrete various stimulatory signals (e.g., vascular endothelial growth factor, interleukin (IL)-8, and TGF-β) that induce MSC homing and subsequent conversion into tumor-associated fibroblasts (TAFs) [42–44]. TAFs promote tumor growth, angiogenesis, and metastasis and suppress anti-tumor immune responses [42, 43]. Irradiated MSCs produce factors, including tumor necrosis factor-α, transforming growth factor-β (TGF-β), IL-6, and fibronectin, implicated in tumor formation and tumor stem cell maintenance [45, 46]. Intra-tumor MSCs may thus survive conventional radiotherapy. Therefore, our understanding of MSC radiobiology also has important implications for understanding cancer progression and for developing new strategies for cancer treatment.

Conclusions

Our study has clearly shown that multiple aspects of the DDR play key roles in promoting the radio-resistance of MSCs. First, we have shown that MSCs retain their capacity to proliferate and to differentiate following IR exposure. Second, while irradiated CD4⁺ CD8⁺ thymocytes fail to execute effective DDR mechanisms and undergo apoptosis, MSCs activate robust DNA damage checkpoints and DNA repair mechanisms that enhance their ability to survive, thereby suppressing apoptosis. Finally, we have shown that MSCs appear to be intrinsically programmed with the ability to effectively execute the DDR following IR treatment by expressing high levels of key DDR proteins such as ATM, Chk2, DNA-PK_cs, and DNA Ligase IV; high levels of the anti-apoptotic Bcl-2 and Bcl-xl; and low levels of pro-apoptotic proteins such as Bim and Puma. MSCs are responsible for supporting and monitoring hematopoiesis, for regenerating various adult tissues, and for modulating immune responses. Therefore, our understanding of how MSCs behave in response to irradiation has direct implications on future advances in improving the success rates of allogeneic BMT, in using MSCs for cellular therapies and in cancer research and treatment.

Acknowledgements

We would like to thank all members of the Ceredig, Lowndes, and Griffin laboratories as well as Dr. Michael D. Rainey for valuable comments and discussions; Michelle M. Duffy, Senthilkumar Alagesan, and Claas Baustian for culturing primary bulk mouse MSCs; Drs. Rainey and Alessandro Natoni for flow cytometry techniques and reagents; Dr. Eva Szegezdi for Caspase-3 and PARP antibodies and Dr. Panagiota Sotiropoulou and Simona Moravcová for providing assistance and reagents for the in vitro DNA DSB end-joining assay. T.S. was supported by an Irish Research Council Government of Ireland Embark Postgraduate Scholarship in Science, Engineering, and Technology (Grant No. RS20102702); Rh.C. by Science Foundation Ireland (Grant No. SFI 09/SRC/B1794) and Stokes Professorship; and N.F.L. and J.A.L.B. by Science Foundation Ireland (Grant No. SFI 07/IN.1/B958). J.A.L.B. is currently affiliated with Systems Biology Ireland, Regenerative Medicine Institute, School of Medicine, Nursing and Health Sciences, National University of Ireland, Galway, Galway, Ireland.

Disclosure of Potential Conflicts of Interest

The authors indicate no potential conflicts of interest.

Author notes

Author contributions: T.S.: conception and design, collection and assembly of data, data analysis and interpretation, and manuscript writing; J.A.L.B.: conception and design, data analysis and interpretation, and manuscript writing; N.F.L. and Rh.C.: conception and design, data analysis and interpretation, manuscript writing, financial support, and final approval of manuscript.

Disclosure of potential conflicts of interest is found at the end of this article.

First published online in STEM CELLSEXPRESS September 7, 2012.