Mitochondrial Superoxide Increases Age-Associated Susceptibility of Human Dermal Fibroblasts to Radiation and Chemotherapy

Cancer incidence and survival rates are expected to increase by over 50% during the next 20 years predominantly due to an aging population and improved therapeutic outcomes (1). By the year 2020, there will be approximately 18,000,000 cancer survivors in the United States (1). Acute and chronic normal tissue injuries resulting from combined modality radiation and chemotherapies represent highly significant treatment limiting complications that are exacerbated in elderly patients (2, 3). Currently, there are limited data available on the age-associated mechanisms of normal tissue injury associated with ionizing radiation (IR) and chemotherapy as well as few treatment options for preventing or mitigating this injury.

Numerous studies examining pathways that connect the processes governing aging and cancer have focused on cellular oxidative metabolism (4–6). Furthermore, normal tissue injury associated with cancer therapy has been hypothesized to involve alterations to mitochondrial oxidative metabolism, leading to increased steady-state levels of reactive oxygen species (ROS), including superoxide (O₂^•−) and hydrogen peroxide (H₂O₂; refs. 7, 8). Recently, several studies have utilized mitochondrially targeted antioxidants to mitigate IR-induced skin damage, preventing dermatitis, inflammation, and fibrosis, thereby preserving the antioxidant capacity of the skin following IR exposure (9, 10). The mitochondrial ETC complexes located on the inner mitochondrial membrane are believed to be a major metabolic site of ROS production in mammalian cells with 0.1% to 1% of the electrons that flow through the ETCs predicted to undergo one-electron reductions of oxygen forming O₂^•− and H₂O₂ (11). Furthermore, increased steady-state levels of ROS are reported to progressively increase with aging (5, 7). Mutations in mitochondrial DNA and nuclear genes encoding mitochondrial proteins are known to accumulate during the aging process (12) and are again associated with increased levels of O₂^•−, H₂O₂, and genomic instability (13, 14). In addition, the age-associated vulnerability of mammalian cells to accumulate DNA damage is also suggested to derive from a progressive decline in DNA repair processes (15). Thus, increased production of O₂^•− and H₂O₂ as well as the accumulation of DNA damage along with changes in mitochondrial function (6, 7, 15) appear to be fundamental to both cancer and aging processes, but the relative contribution of O₂^•− to the increased sensitivity of normal human cells from elderly patients to radiation and chemotherapy is unknown.

The current study demonstrates that exponentially growing early-passage human fibroblasts, from elderly (old) male donor subjects (70, 72, 78 years), are significantly more sensitive to clonogenic killing mediated by platinum-based chemotherapy and IR, relative to young fibroblasts from male donor subjects (5 months and 1 year) and/or adult fibroblasts from a male donor subject (20-year-old). Furthermore, old fibroblasts also showed significant alterations in mitochondrial ETC and aconitase activities. Mitochondrial ETC activity and aconitase activity were restored by SOD2 overexpression, clearly demonstrating a role for mitochondrial O₂^•−in these effects. Donor age-associated alterations in mitochondrial metabolism were found to be accompanied by downregulation of mitochondrial ETC gene expression in old fibroblasts. Furthermore, old fibroblasts also demonstrated elevated levels of endogenous DNA damage that were increased following treatment with IR and chemotherapy. Most importantly, treatment with a small-molecule, superoxide-specific SOD mimetic (GC4419) mitigated the increased sensitivity of old fibroblasts to IR and chemotherapy as well as partially restored mitochondrial function without affecting IR or chemotherapy-induced cancer cell killing. These results show age-associated increases in O₂^•−, DNA-damage, and compromised mitochondrial function result in increased sensitivity of old fibroblasts to IR and chemotherapy that can be inhibited by an SOD mimetic, GC4419, which is currently being studied clinically as a mitigator for oral mucositis in head and neck cancer subjects (NCT02508389). These results support the hypothesis that the use of SOD mimetics in elderly cancer patients undergoing radiation and chemotherapy may have significant benefit in enhancing the tolerability of cancer therapies as well as reducing the chronic toxicities associated with cancer survivorship.

Dermal fibroblasts, from young donor male subjects (5 months and 1 year), adult donor male subject (20-year old) and old donor male subjects (70, 72, 78 years), were obtained from Coriell Institute for Medical Research (Camden, New Jersey). Tissues used to isolate dermal fibroblasts were initially submitted to Coriell from 1973 to 1982. Cell culture identity was authenticated by PCR using microsatellite and Y-chromosome-specific primer pairs. Dermal fibroblasts were obtained from Coriell between 2015 and 2017. The cells were received at passage 12 (Young), passage 13 (Adult) and passages 14 to 15 (Old) from the vendor. All cells were maintained at 37°C in 4% O₂ and 5% CO₂. Mycoplasma was assessed for using the LONZA MycoAlert Mycoplasma Detection Kit. Low-serum fibroblast basal medium (PCS-201-030) from ATCC was used as the cell culture growth medium. The growth medium was supplemented with additional factors in the fibroblasts growth kit-low serum (PCS-201-041). Cells used in all experiments were frozen in the abovementioned media with 10% FBS and 10% DMSO. Cells were passaged using 0.25% trypsin/EDTA. All cells were grown and maintained at 37°C and 4% O₂. Unless specified otherwise, the results were tested in dermal fibroblasts from two young (5-month-old and 1-year-old), one adult (20-year old) and three old (70, 72, and 78 years old) male donors. Data were analyzed by taking the pooled mean of the data from the different cell sources in each age group. Exponentially growing cultures of fibroblasts were used from passages 3 to 7 (passage 1 was defined as when the cells were received from Coriell) for all experiments.

Irradiation was performed on cells plated in 60- or 100-mm dishes with complete media. Typically, cells were allowed to attach for at least 24 to 48 hours at 4% O₂ and 37°C to obtain exponentially growing cultures at around 50%–70% confluence for all experiments unless specified otherwise. Cells were irradiated from 2 to 4 Gy X-rays in the Radiation and Free Radical Research Core laboratory at the University of Iowa. Radiation was given at 1.29 Gy/minute and was delivered using a PANTAK HF-320 ortho volt X-ray unit. All in vitro irradiations used 200 kVp X-rays, 15 mA with a filter composed of 0.35 mm copper (Cu) + 1.5 mm aluminum (Al). After irradiation, the cells were plated for clonogenic cell survival within an hour. For the experiments in vivo, animals were arranged in lead coffins with only their left legs exposed to the X-ray source. All in vivo irradiations used 200 kVp X-rays, 15 mA with a filter composed of 0.35 mm copper (Cu). Radiation was given at 3 Gy/minute with the head height at 30 cm delivered using a PANTAK HF-320 orthovolt X-ray unit. Cisplatin (1 mg/mL; molecular weight 300) was purchased from Hospira, Inc. and carboplatin (10 mg/mL; molecular weight 371) was purchased from APP Pharmaceuticals Inc. (Fresenius Kabi USA) and were further diluted in H₂O prior to use to achieve the desired concentrations. GC4419 (Galera Therapeutics, molecular weight 483) was prepared at a stock concentration of 10 mmol/L in H₂O with 10 mmol/L sodium bicarbonate (pH 7.1–7.4) that was diluted in order to achieve a final concentration of 0.25 μmol/L and was added 72 hours prior to and during the experiment unless specified otherwise. GC4419 (previously known as M40419) is a small molecular weight SOD mimetic specific for dismutation of superoxide with a rate constant of (1 × 10⁷/mol/s; ref. 16), which is comparable with the native SOD enzyme (1.2 × 10⁹/mol/s; ref. 17).

All young (2 months) and adult mice (7 months) were C57BL/6J obtained from The Jackson Laboratory (#000664) and the old (31 months) mice were C57BL/6J obtained from Dr. James Martin's lab at the University of Iowa. All mice were maintained in accordance with ACURF approval #4111207 at the University of Iowa Animal Care facility. Mice were maintained on normal diets and water ad libitum through the course of the experiment. Young, adult, and old mice were irradiated with a single dose of 45 Gy X-rays to the right leg in the Radiation and Free Radical Research Core laboratory at the University of Iowa. Radiation was delivered using a PANTAK HF-320 orthovolt X-ray unit as described above. All animal experiments were conducted in accordance with the University of Iowa Institutional Animal Care and Use Committee.

Growth curves were generated to determine the rate of proliferation in the exponentially growing young and old fibroblasts. Cells were plated in six-well plates with 50,000 cells/dish. Cell numbers were counted days 1 to 5 (using Beckman Z1 particle counter; Beckman Coulter) and the growth curves (cell#/time) were constructed. The doubling time (D_T) was calculated from the exponential part of the growth curve using the equation: D_T = 0.693*t/ ln (N_t/N₀), where t is the time in hours, N_t is number of cells at time t, and N₀ is the number of cells at the initial time. The average D_T is calculated with the doubling times from passages 3 to 7.

To determine the effect of age on the reproductive integrity, young and old fibroblasts were plated at 50,000 per 60-mm dish and 48 hours after plating, both floating and attached cells from control or treated dishes were collected by treatment with trypsin (0.25%). Trypsin was inactivated with full fibroblasts media containing 1% FBS and growth factors following which the cells were centrifuged at 1,200 rpm for 5 minutes and resuspended in fresh media. Cell counts were made with a Coulter Counter and the cells were plated at various densities on lethally irradiated feeder cells (see below) and allowed to grow for 10 to 12 days in complete media at 4% O₂. Subsequently, the cells were stained with Coomassie Blue dye, and the colonies of greater than 50 cells on each plate were counted and recorded, and clonogenic cell survival was determined, as described previously (18). Surviving fraction was determined as the number of colonies per plate divided by the number of cells initially added to that plate. Normalized surviving fraction (NSF) was determined as the surviving fraction of any given clonogenic plate from a treatment group divided by the average surviving fraction of the control (untreated) plates within a given experiment.

During clonogenic survival assays, both young and old fibroblasts were grown on a feeder layer of lethally irradiated Chinese hamster fibroblasts designated as B1 using the complete fibroblast media recipe containing FBS and growth factors. B1 cells (passages 20–40) are maintained in high glucose DMEM 1× media containing L-glutamine and sodium pyruvate and supplemented with 10% fetal bovine serum, 20 mmol/L HEPES, pH 7.3, 1× MEM non-essential amino acids. The B1 feeders were grown at 21% oxygen and the feeder layer was prepared 24 hours prior to any clonogenic survival assay by irradiating the cells with 30 Gy of X-rays at a dose rate of 27 Gy/minute. Each clonogenic experiment had two extra dishes plated with feeder cells to ensure no colony growth at the end of the 10- to 12-day incubation.

Steady-state levels of superoxide were estimated using oxidation of the fluorescent dye, dihydroethidium (DHE) purchased from Invitrogen as described previously (13). Following labeling with 10 μmol/L DHE at 37°C for 20 to 30 minutes, the samples were analyzed using LSR violet flow cytometer (Becton Dickinson Immuno-cytometry System, Inc.; excitation 405 and 488 nm, emission 585 nm band-pass filter). The mean fluorescence intensity (MFI) of 10,000 cells was analyzed per sample and corrected for auto-fluorescence from unlabeled cells. Each sample was normalized to the young fibroblast group, to yield the normalized mean fluorescent intensity (NMFI).

Estimates of steady-state levels of mitochondrial superoxide in exponentially growing young and old fibroblasts were assessed using the cationic dye MitoSOX red (Molecular probes). Following labeling with 2 μmol/L MitoSOX red, the cells were analyzed using LSR violet flow cytometer (Becton-Dickinson; excitation 488 nm, emission 585 nm bandpass filter). The MFI of 10,000 cells was analyzed per sample and corrected for autofluorescence from unlabeled cells. Each sample was normalized to the young fibroblasts to yield the NMFI.

Membrane potential of young and old fibroblasts was detected by JC-1 fluorescence. JC-1 was used at a final concentration of 3 μmol/L. The mitochondrial ETC uncoupler CCCP (Carbonyl cyanide m-chloro phenyl hydrazone) was used at 50 μmol/L as the negative control. The red and green fluorescence of 10,000 cells was recorded per sample, and the mean of three samples was calculated for each condition. Each sample was normalized to the young fibroblast group to yield the normalized red/green ratio.

Measurements of the ETC complex activities were done as described previously (19, 20). Whole cell pellets from exponentially growing cultures were disrupted by sonication and freeze thaw. The assays were performed on a Beckman DU 800 spectrophotometer. Complex I activity was assayed as the rate of rotenone-inhibitable NADH oxidation as described previously (19, 20). Complex II activity was assayed as the rate of reduction of 2, 6-dichloroindophenol by coenzyme Q in the presence and absence of 0.2 M succinate as described previously (19, 20). Complex III activity was assayed as the rate of cytochrome c reduction by coenzyme Q2 whereas, complex IV activity was assayed as the rate of cytochrome c oxidation (19, 20). Complex V activity assay was performed at 340 nm. The activity was measured using lactate dehydrogenase and pyruvate kinase as the coupling enzymes. All the ETC complex activities were normalized to citrate synthase activity for mitochondrial content.

Cellular citrate synthase activity was measured using the protocol previously described (21). Protein was quantified using the Lowry assay. Fifty micrograms protein was combined with NADP+ oxaloacetic acid (0.5 mmol/L) and dTNB (0.1 mmol/L). Reactions were monitored at 412 nm for the formation of TNB formed every 30 minutes for 2 minutes. Rates were determined from slopes determined by regression analysis of the data.

Total cellular aconitase activity was measured using the protocol previously described (22). Protein was quantified using the Lowry assay (23). Rates were determined from slopes determined by regression analysis of data.

Total RNA was isolated from exponentially growing young (5-month-old donor) and old (78-year-old donor) fibroblasts using the RNeasy mini kit from Qiagen (Cat # 74104). Briefly, cells were trypsinized with 0.25% Trypsin-EDTA, washed in DPBS, and pelleted. The RNA was isolated using the manufacturer's instructions. Total RNA concentration and purity in the eluted samples was quantified using a NanoDrop ND-100 UV spectrophotometer (NanoDrop Technologies).

The extracted RNA was converted to cDNA using the RT² HT First Strand Kit from Qiagen (cat # 330401). The cDNA was stored at −20°C until used for real-time PCR profiling using the mitochondrial energy metabolism PCR array from Qiagen (# PAHS008Z).

Differences in the nuclear-encoded genes of the mitochondrial ETC were determined using the RT² Profiler PCR Array from Qiagen (Cat # PAHS008Z; accession number: GSE100921). Relative gene expression changes and fold changes were determined using the ΔΔCt method as per the manufacturer's instructions. Fold change was quantified as the normalized gene expression in the old fibroblasts (ΔΔCt^old) divided by the normalized gene expression in the young fibroblasts (ΔΔCt^young). Fold-change values greater than two were considered positive or upregulation, whereas fold-change values less than −2 indicated negative or downregulation in gene expression.

Twenty-five nanograms cDNA samples from young and old fibroblasts was used to perform the qPCR analyses of the 13 mitochondrially encoded genes of the ETC. Reactions were performed in triplicate for each of the genes. All primers were purchased from Applied Biosystems including the two ribosomal RNAs (RNR1 and 2). qPCR cycle conditions were 2 minutes at 50°C, 10 minutes at 95°C followed by 40 cycles of 15 seconds of denaturation at 95°C and 60 seconds of annealing/extension at 60°C. Fluorescent signal intensity of PCR products was recorded and analyzed on ABI-7900 (Applied Biosystems) using the SDS 2.3 software. The threshold cycle or Ct value within the linear exponential phase of the curve was used to calculate the ΔΔCt value. The RNR2 encoded in the mitochondrial DNA was used as housekeeping gene. Fold change was quantified as the normalized gene expression in fibroblasts from the 78-year-old donor (ΔΔCt^old) divided by the normalized gene expression in fibroblasts from the 5-month-old donor (ΔΔCt^young). Fold-change values greater than 1 were considered positive or upregulation, whereas fold-change values less than −1 indicated negative or downregulation in gene expression.

Relative levels of DNA damage were determined using γH2AX staining with both confocal microscopy and flow cytometry (24). For the flow cytometric analysis, briefly, cells were trypsinized, washed in PBS, and fixed overnight in 70% ethanol. Following day, the cells were resuspended in cold PBST (PBS-tween) to rehydrate on ice for 10 minutes, centrifuged at 1,200 rpm for 5 minutes at 4°C, and resuspended overnight in rabbit polyclonal anti-phospho H2AX antibody at 1:800 in PBST purchased from Cell signaling technology (Cat # 2577). The cells were then washed with 2% FBS in PBS and resuspended in a FITC-conjugated goat anti-rabbit secondary antibody at 1:200 in PBST for an hour at room temperature. Cells were then rinsed with PBS, centrifuged at 1,200 rpm for 5 minutes at 4°C, resuspended in fresh PBS and analyzed by flow cytometry using LSR-violet. The MFI of 10,000 cells was analyzed per sample and corrected for autofluorescence from unlabeled cells. Each sample was normalized to the control young fibroblasts to yield the NMFI.

For analysis by confocal microscopy, cells were plated on Thermo Scientific Nunc Lab-Tek II Chamber Slide System from Thermo Fisher Scientific(Cat # 154453PK) at least 24 hours before the experiment. Cells were fixed in 4% PFA (paraformaldehyde) for 15 minutes at 37°C, washed with PBS and blocked in 1% donkey serum in TBST for 30 minutes at room temperature. Cells were then treated with a mouse monoclonal anti-phospho H2AX antibody at 1:1,000 in TBST purchased from Millipore (Cat # 05-636-I) overnight. The following day, the chambers were washed three times with PBST and resuspended in a FITC-conjugated goat anti-mouse secondary antibody at 1:200 in PBST for an hour at room temperature for 2 hours. The samples were then washed three times with TBST and confocal analysis was performed. Hoechst was used as a nuclear stain at 1:1,000. Samples were scored in a blinded-fashion and an average of at least 150 nuclei was scored from at least 4 randomly selected areas for each group. The fluorescence intensities were quantified using the ImageJ software.

SOD activity was determined using the indirect competitive inhibition assay as described previously (25). Briefly, the flux of superoxide generated by the reaction of xanthine & xanthine oxidase was measured using nitroblue tetrazolium (NBT). Increasing concentrations of SOD rapidly converts the superoxide produced to hydrogen peroxide and the rate of NBT reduction is inhibited and can be measured spectrophotometrically at 560 nm. The rate of NBT reduction is calculated as percent inhibition, relative to the NBT reduction in the absence of the sample and the amount of protein causing 50% maximum inhibition is defined as one unit of activity. Exponentially growing young and old fibroblasts were washed and scraped in ice cold PBS and spun down. The dry pellets were then resuspended in DETAPAC buffer (0.05 M phosphate buffer, pH 7.8, 1.34 mmol/L DETAPAC) and MnSOD activity was distinguished from CuZSOD activity using 5 mmol/L NaCN as described (25).

Exponentially growing young and old fibroblasts were fixed in 4% PFA for 15 minutes at 37°C. Cells were washed 3× with TTBS (tris buffered saline containing tween) and then blocked for 1 hour at room temperature in 4% normal donkey serum and TTBS. Cells were then washed 3× in TTBS and incubated in primary antibody (Rb α-TOM20—an outer mitochondrial membrane protein) overnight at 4°C. The following day, cells were washed 3× in TTBS and incubated with secondary antibody (dnky α-Rb 488) for 1 hour at room temperature. Cells were incubated in PBS containing Hoechst 33342 (1:1,000, 1 μg/mL) for 5 minutes at room temperature and then washed 3× with PBS before imaging.

All groups were imaged in a blinded fashion on a Leica DMI4000B epifluorescence microscope with a 100× oil immersion imaging approximately 50 to 100 cells/condition. Images were preprocessed and mitochondrial morphology was analyzed using an ImageJ macro as previously described (26, 27). Analyses of the mitochondrial morphology were carried out by an observer blinded to conditions. Shape metrics including form factor (perimeter²/(4 × π × area) and aspect ratio (long axis/short axis of a best fit ellipse) were determined in all groups. Both the shape metrics have a minimum value of 1 for perfect circular mitochondria. All groups were normalized to the young fibroblast control group.

Skin samples were collected from both IR exposed and control legs of young, adult, and old mice through the University of Iowa Comparative Pathology Core. Each skin tissue was equally divided into two parts, one half of which was fixed in formalin while the other half was frozen in Tissue-Tek OCT. Compound (Sakura Finetek, VWR). The formalin-fixed tissue sections were stained with hematoxylin and eosin (H&E) for determining changes in epidermal and dermal thickness, whereas the samples frozen in OCT. were utilized for DHE staining.

From OCT-frozen mouse skin tissue, 10-μm sections IR exposed skin was cut and placed on the same slide to control skin for that specific group (e.g., young control vs. young IR). For comparison between different age groups, sections from the same treatment group but different age groups were placed on the same slide (e.g., young IR vs. adult IR) in order to maintain similar experimental labeling conditions. Tissue sections were stained with 10 μmol/L DHE for 10 minutes at 37°C in PBS containing 5 mmol/L sodium pyruvate prior to analysis by confocal microscopy using the Olympus Fluoview FV1000 confocal microscope. All sections were imaged at 20× magnification, 37°C using Cy3 as the fluorochrome. For treatment groups, tissue sections were pretreated for 30 minutes with 0.25 μmol/L GC4419 SOD mimetic or, alternatively, as a positive control, tissue sections were treated with 10 μmol/L antimycin A (AntA) and DHE for 10 minutes. Each image obtained was quantified by measuring the MFI of at least 150 cell nuclei, and normalized to young mouse control or the respective young treatment group.

Skin thickness was measured in the tissue slices from young, adult, and old mice both from the IR-exposed and control leg. Sections stained with H&E were used to determine differences in epidermal and dermal layers of the skin. Images were generated using the Olympus BX-61 motorized light microscope at 20× magnification and were scored in a blinded fashion. Epidermal and dermal layers were measured using the arbitrary line function, which is integrated in the imaging software Cellsens on the same microscope. At least 30 different (randomly selected) areas were sampled for changes in the epidermal and dermal layers of the skin per treatment group.

Oxygen consumption was measured using the Clark electrode in young and old fibroblasts as described previously (28). Exponentially growing young and old fibroblasts were plated at 200,000/dish, grown 48 hours, trypsinized, and resuspended in PBS at 37°C and O₂ consumption monitored for 30 minutes using the Clark electrode.

ATP production was determined using a kit (Promega, CellTiterGlo) by measuring luminescence on a SpectraMax Microplate reader. Briefly, a cell suspension made during exponential growth phase (50,000 cells/100 μL) was used for the kit-based assay. Instructions were followed according to the manufacturer's protocol. A standard curve of ATP concentration versus luminescence signal was generated with adenosine 5′-triphosphate (ATP) disodium salt hydrate (CAS # 34369-07-8, Sigma-Aldrich).

Statistical analyses were done using the graph pad prism software. Statistical significance was determined using one-way ANOVA for non-parametric measurements. Tukey post hoc test was performed wherever appropriate. Unless specified otherwise in the figure legend, error bars represent ± SEM and statistical significance was defined as **, P < 0.05.

Exponentially growing early-passage old fibroblasts demonstrate significantly increased levels of DHE and MitoSOX oxidation relative to young and adult fibroblasts (Fig. 1A and B). In order to confirm that the increased DHE and MitoSOX oxidation in old fibroblasts is caused by an increase in endogenous steady-state levels of O₂^•−, we utilized the small molecule superoxide-specific SOD mimetic, GC4419, developed by Galera Therapeutics Inc. (Supplementary Fig. S1A) to inhibit the oxidation of DHE or MitoSOX (16). In the presence of 0.25 μmol/L SOD mimetic, old fibroblasts showed a 4-fold increase in GC4419-inhibitable DHE oxidation (Fig. 1C) and a 3-fold increase in GC4419-inhibitable MitoSOX oxidation (Fig. 1D) clearly demonstrating significantly increased steady-state levels of both cytosolic and mitochondrial superoxide, relative to both young and adult fibroblasts.

In vitro, old fibroblasts relative to young fibroblasts demonstrate significantly increased doubling times (Supplementary Fig. S1B), decreased ability to undergo population doublings (Supplementary Fig. S1C), and decreased plating efficiencies (Supplementary Fig. S1D). To determine if there is a difference between young and old fibroblasts in sensitivity to radiation and platinum-based chemotherapy, exponentially growing fibroblasts from young donor male subjects and old donor male subjects were treated with increasing doses of IR or 24-hour exposures to platinum-based chemotherapy and assessed for clonogenic survival. Results indicate that clonogenic survival was significantly reduced in the old fibroblasts relative to the young fibroblasts following treatment with IR or platinum-based chemotherapy (Supplementary Fig. S1E–S1G).

When early passage, exponentially growing young, adult, and old fibroblasts were exposed to GC4419 (0.25 μmol/L) 24 hours prior to, during, as well as following treatment (in the cloning dishes) with IR and platinum-based chemotherapy, GC4419 was able to significantly protect old fibroblasts, from clonogenic cell killing (Fig. 1E and F). Interestingly, no significant differences in clonogenic cell killing were observed in the GC4419-treated young and adult fibroblasts following IR, cisplatin and/or carboplatin treatment, relative to similarly treated cells in the absence of GC4419 (Fig. 1E and F).

In order to determine if increasing mitochondrial superoxide could sensitize young fibroblasts to cisplatin, carboplatin, or radiation, young fibroblasts were treated with the ETC blocker, rotenone, which is known to increase steady-state levels of mitochondrial superoxide (29, 30). Treatment of young fibroblasts with 0.5 μmol/L rotenone for 24 hours led to increased steady-state levels of superoxide as determined by DHE oxidation (Fig. 1G) as well as sensitizing the young fibroblasts to clonogenic cell killing by cisplatin, carboplatin, or IR (Fig. 1H). This rotenone-induced sensitization to IR and chemotherapy was mitigated when the young fibroblasts were treated with 0.25 μmol/L GC4419 before, during, and following exposure to rotenone, IR and the platinum-based chemotherapy drugs (Fig. 1H). These results continue to support the conclusion that young fibroblasts can be sensitized to IR and platinum-based chemotherapy by increasing mitochondrial superoxide.

Most importantly, treatment with 0.25 μmol/L GC4419 before, during, and following treatment with IR and platinum-induced killing did not protect exponentially growing human non–small cell lung cancer (H292) or head and neck cancer cells (SCC25; Fig. 1I and J) from radiation-and platinum-based chemotherapy toxicity. The ability of GC4419 to selectively protect normal cells from IR and chemotherapy while maintaining tumor cell killing is critical to translating these results into the clinical setting.

Several studies have linked platinum-based chemotherapy agents and IR to DNA damage (31, 32). In addition, there is a strong mechanistic connection between increased steady-state levels of ROS and DNA damage (8). Consistent with the role of DNA damage in normal tissue injury, exponentially growing old fibroblasts demonstrate increased fluorescence associated with immunoreactive γH2AX staining, relative to young fibroblasts that were exacerbated following exposure to IR (1 hour) or platinum-based chemotherapy (24 hour; Fig. 2A and B). This increase in fluorescence associated with immunoreactive γH2AX staining was partially inhibited by GC4419 (Fig. 2A). These data strongly support the hypothesis that increased sensitivity to IR and platinum-based chemotherapy in the old fibroblasts is causally related to increased steady-state levels of O₂^•− enhancing residual DNA damage. Importantly, GC4419 did not interfere in vitro (Fig. 1I and J) or in vivo (33), with the anticancer activities of IR or platinum-based chemotherapy.

Mitochondrial ETCs are a major site of O₂^•− production and disruption of the flow of electrons through ETCs is associated with increased steady-state levels of O₂^•− (13, 14, 34, 35). As old fibroblasts have increased steady-state levels of mitochondrial O₂^•− (Fig. 1A and B), we hypothesized that old fibroblasts would also demonstrate alterations in ETC complex activities. Whole cell homogenates from exponentially growing old fibroblasts demonstrated significantly reduced activities of complexes I, II, IV, V (Fig. 3A). Total aconitase activity (normalized to citrate synthase activity) was also significantly decreased in both adult and old fibroblasts relative to young fibroblasts (Fig. 3B). In contrast, there was no significant change in citrate synthase activity, which is a Krebs cycle enzyme used to assess total mitochondrial content (Fig. 3C; ref. 21). Consistent with decreased mitochondrial ETC complex activities and increased levels of O₂^•− in mitochondria, there was also an increase in mitochondrial membrane potential determined by JC-1 (Fig. 3D), an increase in oxygen consumption (Fig. 3E) and a decrease in steady-state levels of ATP (Fig. 3F) in exponentially growing old fibroblasts, relative to young fibroblasts. These results support the hypothesis that old fibroblasts exhibit significant reductions in mitochondrial metabolic efficiency. Taken together, the results in Figs. 1 to 3 indicate that old fibroblasts (relative to young fibroblasts) demonstrate significant alterations in mitochondrial oxidative metabolism that are accompanied by increased steady-state levels of O₂^•−, which contribute to differential sensitivity to IR and chemotherapy.

Significant age-associated mitochondrial elongation was also noted in exponentially growing old fibroblasts, relative to the young fibroblasts (Fig. 4A). Both mitochondrial form factor (a measure for mitochondrial elongation with a minimum value of 1 for perfectly round mitochondria; Fig. 4B) and aspect ratio (aspect ratio = major axis/minor axis; Fig. 4C) were significantly increased in old fibroblasts relative to young fibroblasts. Furthermore, treatment with GC4419 significantly decreased mitochondrial elongation in old fibroblasts, making them comparable with young fibroblasts, indicating that increased steady-state levels of O₂^•− mediate the age-associated changes in mitochondrial structure (Fig. 4A–C). In contrast, MitoTracker green levels (a measure of mitochondrial mass) remained unchanged between young and old fibroblasts, suggesting no change in overall mitochondrial mass (Fig. 4D). These data support the hypothesis that old fibroblasts undergo changes in mitochondrial morphology to maintain cellular homeostasis and function as well as that O₂^•− is a mediator of donor age-associated mitochondrial elongation.

In addition to changes in mitochondrial morphology, exponentially growing fibroblasts from elderly donor patients also showed changes in the stoichiometry of ETC complex gene expression as demonstrated by qPCR arrays performed on both nuclear as well as mitochondrial-encoded genes comparing young (5-month-old donor) and old (78-year-old donor) fibroblasts. Relative to young fibroblasts (5-month-old donor), the old fibroblasts (78-year-old donor) demonstrated an age-associated ≥2-fold change in expression of 48 nuclear encoded ETC genes (Fig. 5A). In contrast, there were only 12 genes that showed a ≥2-fold change in the nuclear encoded ETC gene expression between two different young fibroblasts (comparing 5-month-old and 1-year-old donors; Fig. 5B). Interestingly, pretreatment of exponentially growing young and old fibroblasts with GC4419 was able to reduce the number of nuclear genes that demonstrated ≥2-fold changes in expression from 48 to 12 (Fig. 5C). Consistent with changes in nuclear encoded ETC genes, old fibroblasts (78-year-old donor) demonstrated an age-associated ≥2-fold change in expression of two mitochondrial encoded ETC genes (Fig. 5D) while the mitochondrial gene expression between fibroblasts from two young donors (comparing 5 months old with 1 year old) showed no ≥2 fold changes (Fig. 5E). Furthermore, pretreatment of exponentially growing young and old fibroblasts with GC4419 showed no ≥2-fold changes in expression of mitochondrial encoded ETC genes (Fig. 5F).

Taken together, the results in Figs. 4 and 5 indicate that structural changes as well as alterations in the stoichiometry of ETC gene expression occur in old fibroblasts (relative to young fibroblasts) that could be related to differences in mitochondrial function and may also be related to increased steady-state levels of O₂^•−.

To determine if increased steady-state levels of O₂^•− could be linked to age-associated changes in mitochondrial metabolism, old fibroblasts were treated with the SOD mimetic, GC4419, or transduced to increase mitochondrial manganese superoxide dismutase activity (aka, SOD2) using an adenoviral vector (Ad-MnSOD) followed by analysis of aconitase, complex I, and complex V activity (Fig. 6A–C). Interestingly, basal levels of total SOD and CuZnSOD activity showed a significant age-associated decrease in the adult and old fibroblasts relative to the young fibroblasts (Fig. 6D). However, there was no significant difference in CuZnSOD activity of old fibroblasts treated with either 50 MOI Ad-Empty or Ad-MnSOD (Fig. 6E). Treatment of young, adult, and old fibroblasts with 50 MOI Ad-MnSOD or 0.25 μmol/L GC4419 resulted in significantly increased cyanide-resistant SOD activity, relative to Ad-Empty vector or vehicle control (Fig. 6F). Both in the adult and old fibroblasts (but not in the young fibroblasts), the increase in cyanide-resistant SOD activity was able to significantly increase the activity of aconitase relative to vector control (Fig. 6A). In addition, overexpression of MnSOD also significantly increased ETC complex I (Fig. 6B) and complex V (Fig. 6C) activities, relative to vector control. Likewise, treatment of old fibroblasts with SOD mimic (GC4419) also significantly increased the activity of aconitase (Fig. 6A), ETC complex I (Fig. 6B), and complex V (Fig. 6C), relative to these activities in vehicle control fibroblasts. Furthermore, the SOD activity of GC4419 was retained in full media and in DPBS for 2 weeks at 37°C in a tissue culture incubator (Fig. 6G) demonstrating the stability of SOD activity in these model systems. Taken together, these results provide strong evidence for the hypothesis that increases in O₂^•− mediate the age-associated alterations in structure as well as function of mitochondrial ETC and TCA cycle proteins.

A previously described (36) model of IR-induced cutaneous right hind limb skin injury induced by a single fraction of 45 Gy was delivered to four young (2 months), three adult (7 months), and three old (31 months) C57BL/6J male mice. Skin samples from the irradiated and contralateral unexposed leg were collected 6 weeks following IR and analyzed for fibrosis as well as oxidative stress markers. Two of the three old mice died within 2 weeks of IR treatment, while none of the young or adult mice demonstrated mortality. There were no changes in behavior or feeding observed in the 6 weeks following IR exposure in the surviving animals. Skin samples from both adult and old mice showed an increase in basal levels of DHE oxidation relative to young mice (Fig. 7A) in both the dermis (Fig. 7B) and epidermis (Fig. 7C). Furthermore, these increases in basal DHE oxidation in the adult and old mouse skin were significantly attenuated when the sections were treated with GC4419 prior to DHE staining, both in the dermis (Fig. 7B) and the epidermis (Fig. 7C) compared with the young mice. In addition, skin samples from irradiated adult and old mice also showed a significant 2- to 3-fold increase in DHE oxidation levels in comparison with skin samples from irradiated young mice (Fig. 7A) in both the dermis and epidermis (Fig. 7B and C), the majority of which were inhibited by GC4419, suggesting a causal role for O₂^•− in the changes in DHE oxidation. No significant change in DHE oxidation was observed in skin samples from the irradiated leg of young mice relative to the skin samples from the unirradiated leg of the same mice. The increases in DHE oxidation at baseline and following IR exposure in the adult and old mice were accompanied by a significant increase in IR-induced epidermal thickening (Fig. 7D and E). These results, combined with the in vitro data (Fig. 1), support the hypothesis that changes in steady-state levels of O₂^•− could contribute to age-associated skin damage and fibrosis following IR exposure.

Progressive deterioration of mitochondrial function and structure leading to increased steady-state levels of superoxide have been implicated in the degenerative processes associated with aging and onset of cancer as well as the persistent normal tissue injury processes initiated by radiation exposure (7, 37–40). However, clear demonstrations of specific changes in mitochondrial function and structure leading to the mechanisms underlying increased levels of superoxide and how these changes might contribute to the enhanced normal tissue injury seen in older patients treated with radiation and chemotherapy are lacking (2, 3). In addition, mitochondrial fidelity proteins including the sirtuins and several cell-cycle regulatory proteins including p21, p53, and p16^INK4A have been shown to govern the process of aging via a mechanism that involves modulation of ROS levels (6, 41–43).

In the current study, we demonstrate in exponentially growing normal human dermal fibroblasts that age-associated changes in mitochondrial structure, function, and ETC complex activity as well as differential sensitivity to IR and chemotherapy are mediated by increased steady-state levels of mitochondrial O₂^•−. It has been suggested that aging-associated changes in the stoichiometry of ETC protein complexes could change the residence time of electrons at sites accessible to O₂, leading to increased one electron reductions of O₂ to form superoxide (7). The data in Figs. 3 through 6 clearly show on a global level that the stoichiometry of both the activities and expression of genes coding for ETC proteins demonstrate age-associated changes that could potentially contribute to increased steady-state levels of superoxide as well as the structural and functional changes seen in mitochondria.

Advanced age is a clinical factor associated with the severity of radiation-induced skin reactions independent of total radiation dose or dose rate (44). In addition, age-associated increased sensitivity to radiation in the skin has been proposed to be related to decreased epidermal turnover, mitosis, malnutrition, or obesity (45) and thus may not be related to the phenomenon being described in this manuscript.

A detailed mechanistic understanding of the role that metabolic production of O₂^•− plays in the differential toxicity seen in elderly human cells exposed to chemo-radiation could provide a clear biochemical rationale for the development of agents that protect normal tissues from elderly patients during cancer therapy. Using this rationale, we predicted that reducing O₂^•− levels and maintaining the integrity of DNA in the face of metabolic oxidative stress induced by chemo-radiation could enhance protection of normal human dermal fibroblasts from older subjects. Previous studies have suggested age-associated increased levels of oxidative damage to proteins (46–48) and DNA could potentially contribute to sensitivity to therapy in elderly cancer patients (12, 49) but no direct invovlement of superoxide or mitochondrial oxidative metabolism in these processes has been demonstrated. The current study extends these observations and provides strong proof-of-concept support for the hypothesis that age-associated differences in mitochondrial oxidative metabolism leading to increased steady-state levels of O₂^•− mediate the age-associated differential susceptibility of human dermal fibroblasts to IR and chemotherapy. Furthermore, the current study also shows that a small molecule SOD mimetic, GC4419, is able to restore metabolic activity and reduce susceptibility to platinum-based chemotherapy, as well as IR injury in fibroblasts from elderly subjects relative to those from young subjects. GC4419 was also able to significantly ameliorate the IR-induced increases in steady-state levels of O₂^•− in aging skin both in the presence and absence of IR exposure. Importantly, GC4419 did not inhibit human cancer cell killing mediated by IR or platinum-based chemotherapies. The current study supports the hypothesis that a clear mechanistic understanding of the relationship between mitochondrial O₂^•-− metabolism, aging, and cancer therapy response outcomes may provide a useful tool for the protection of normal tissues during combined modality therapies in elderly cancer patients.

K.A. Mapuskar reports receiving a commercial research grant from Galera Therapeutics. V. Monga reports receiving commercial research grants from Orbus Therapeutics and Immunocellular Therapeutics. D.R. Spitz reports receiving a commercial research grant from Galera Therapeutics and is a consultant/advisory board member for Galera Therapeutics. B.G. Allen reports receiving a commercial research grant from Galera Therapeutics Inc. and is a consultant/advisory board member for Galera Therapeutics. No potential conflicts of interest were disclosed by the other authors.

Conception and design: K.A. Mapuskar, J.D. Schoenfeld, T.A. Hejleh, M. Furqan, F.E. Domann, D.R. Spitz, B.G. Allen

Development of methodology: K.A. Mapuskar, S. Strack, T.A. Hejleh, M. Furqan, F.E. Domann, P.C. Goswami, D.R. Spitz, B.G. Allen

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): K.A. Mapuskar, K.H. Flippo, J.D. Schoenfeld, M. Furqan, V. Monga, P.C. Goswami, D.R. Spitz, B.G. Allen

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): K.A. Mapuskar, T.A. Hejleh, J.M. Buatti, P.C. Goswami, D.R. Spitz, B.G. Allen

Writing, review, and/or revision of the manuscript: K.A. Mapuskar, K.H. Flippo, D.P. Riley, T.A. Hejleh, M. Furqan, V. Monga, F.E. Domann, J.M. Buatti, D.R. Spitz, B.G. Allen

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): K.A. Mapuskar, D.P. Riley, J.M. Buatti, D.R. Spitz, B.G. Allen

Study supervision: K.A. Mapuskar, S. Strack, D.R. Spitz, B.G. Allen

Other (provided experimental compounds): D.P. Riley

We thank the Flow Cytometry Facility as well as the Radiation and Free Radical Research Core for support of these studies. The authors would like to thank Amanda Kalen and Dr. Michael L. McCormick from the Radiation and Free Radical Research Core at the University of Iowa for assistance with irradiations and spectrophotometric assays respectively. The authors would also like to thank Katherine N. Gibson-Corley from the Comparative Pathology Laboratory (CPL) in the Department of Pathology at the University of Iowa for her assistance in processing skin samples from untreated and irradiated mice. The authors would also like to thank Dr. Zita Sibenaller and Timothy Waldron for assistance in the lab and mouse irradiation experiments, respectively. The authors would also like to thank Brett Wagner and Dr. Claire Doskey for helpful advice with Clark electrode and ATP assays, respectively. The authors thank Dr. Robert Beardsley and Dr. Jeffery Keene, of Galera Therapeutics, Inc., for their thoughtful review and comments on this manuscript and Gareth Smith for assistance with graphics.

This research was also supported in part by the NIH grant P30CA086862 (D.R. Spitz) as well as R01CA111365 (P.C. Goswami), R01NS056244 (S. Strack), T32 CA078586 (D.R. Spitz), and R01CA182804 (D.R. Spitz). This work was also supported in part by ASTRO JF2014-1 (B.G. Allen), a Medical Research Initiative Grant from the Carver Trust (B.G. Allen), The Carver Research Program of Excellence in Redox Biology and Medicine (D.R. Spitz), and the Department of Radiation Oncology at the University of Iowa.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.