The Role of Metabolic Flexibility in the Regulation of the DNA Damage Response by Nitric Oxide

In this report, we show that nitric oxide suppresses DNA damage response (DDR) signaling in the pancreatic β-cell line INS 832/13 and rat islets by inhibiting intermediary metabolism. Nitric oxide is known to inhibit complex IV of the electron transport chain and aconitase of the Krebs cycle.

ABSTRACT

In this report, we show that nitric oxide suppresses DNA damage response (DDR) signaling in the pancreatic β-cell line INS 832/13 and rat islets by inhibiting intermediary metabolism. Nitric oxide is known to inhibit complex IV of the electron transport chain and aconitase of the Krebs cycle. Non-β cells compensate by increasing glycolytic metabolism to maintain ATP levels; however, β cells lack this metabolic flexibility, resulting in a nitric oxide-dependent decrease in ATP and NAD⁺. Like nitric oxide, mitochondrial toxins inhibit DDR signaling in β cells by a mechanism that is associated with a decrease in ATP. Non-β cells compensate for the effects of mitochondrial toxins with an adaptive shift to glycolytic ATP generation that allows for DDR signaling. Forcing non-β cells to derive ATP via mitochondrial respiration (replacing glucose with galactose in the medium) and glucose deprivation sensitizes these cells to nitric oxide-mediated inhibition of DDR signaling. These findings indicate that metabolic flexibility is necessary to maintain DDR signaling under conditions in which mitochondrial oxidative metabolism is inhibited and support the inhibition of oxidative metabolism (decreased ATP) as one protective mechanism by which nitric oxide attenuates DDR-dependent β-cell apoptosis.

INTRODUCTION

Proinflammatory cytokines, such as interleukin-1β (IL-1β), interferon gamma (IFN-γ), and tumor necrosis factor alpha (TNF-α), that are released in and around islets during insulitis are thought to contribute the loss of functional β-cell mass during the pathogenesis of type 1 diabetes (1–3). Nitric oxide, produced at micromolar levels by β cells following cytokine-stimulated expression of the inducible isoform of nitric oxide synthase (iNOS), mediates the damaging actions of cytokines on β cells (4–6). This damage includes the inhibition of insulin secretion and mitochondrial oxidative metabolism (5, 7), induction of endoplasmic reticulum (ER) stress (8–10), inhibition of protein synthesis (11), and damage to DNA (12–14). While damaging, nitric oxide also activates a number of protective pathways that support β-cell survival (15). In addition to stimulating the expression of genes involved in base excision repair of damaged DNA (16), nitric oxide also regulates the DNA damage response (DDR) in β cells (17, 18). The DDR participates in DNA repair as well as the coordination of cell fate decisions in response to DNA double-strand breaks (DSBs) (19, 20). We have shown that prolonged exposure of rat islets and insulinoma cells to cytokines results in the activation of ataxia telangiectasia mutated (ATM), a primary signal-transducing kinase of the DDR (21), and ATM-dependent apoptosis of β cells (18). True to its Janus nature, nitric oxide is responsible for cytokine-induced DDR activation via the induction of DNA strand breaks, yet it also limits DDR signaling and protects β cells from DNA damage-dependent apoptosis (17). The inhibition of the DDR by nitric oxide occurs under conditions in which β cells are actively producing this free radical, and the response is selective for β cells, as we have yet to observe it in any other cell type examined to date (17, 22).

The goal of the present study was to determine the mechanisms by which nitric oxide selectively attenuates DDR signaling in pancreatic β cells. Unique to β cells is the coupling of glycolytic and mitochondrial oxidation, where 90% of the carbons of glucose are oxidized to CO₂, and this oxidation occurs on substrate supply (23, 24). It is the oxidation of glucose in this manner that allows the β cell to sense blood glucose levels and provide the appropriate amount of insulin to remove glucose from the bloodstream (25). Glucose oxidation results in the accumulation of ATP (increase in the ATP/ADP ratio) and the closure of ATP-sensitive K_ATP channels, resulting in β-cell depolarization, calcium entry, and calcium-induced exocytosis of insulin granules (25). In non-β cells, glycolysis and mitochondrial oxidation are not coupled in this manner (26), allowing these cells to maintain ATP levels in the presence of inhibitors of mitochondrial oxidation by increasing glycolytic flux. In this report, we show that the inhibitory actions of nitric oxide on DDR signaling in β cells are associated with an inhibition of intermediary metabolism. β cells are unable to compensate for impaired mitochondrial respiration with an increase in glycolytic flux, resulting in decreased levels of NAD⁺ and ATP (23). Non-β cells compensate for nitric oxide-mediated inhibition of mitochondrial oxidation with enhanced glycolysis, allowing these cells to maintain ATP levels and DDR signaling. Consistent with the effects of nitric oxide, inhibitors of mitochondrial oxidative metabolism attenuate DDR signaling selectively in β cells, and forcing non-β cells to rely on mitochondrial oxidative metabolism for ATP generation sensitizes these cells to DDR inhibition by nitric oxide and mitochondrial toxins. These findings suggest that the selective inhibition of DDR signaling by nitric oxide in β cells is associated with the coupling of glycolytic and mitochondrial oxidative metabolism and that non-β cells maintain DDR signaling due to their metabolic flexibility in shifting to glycolytic metabolism for ATP generation when mitochondrial respiration is impaired.

RESULTS

Nitric oxide-dependent inhibition of γH2AX correlates with a decrease in cellular levels of ATP.

We have previously shown that nitric oxide inhibits DDR signaling selectively in insulin-containing β cells (17, 22). Consistent with these previous findings, the nitric oxide donor (Z)-1-[N-(3-aminopropyl)-N-(3-ammoniopropylamino)]diazen-1-ium-1,2-diolate (DPTA/NO) inhibits camptothecin-induced DDR signaling (assessed by γH2AX formation) in INS 832/13 cells and H₂O₂-induced γH2AX formation in rat islet cells (Fig. 1A and C). The inhibitory actions of DPTA/NO are concentration dependent and correlate with a decrease in the cellular levels of ATP (Fig. 1B and D). At a concentration of 200 μM, DPTA/NO attenuates γH2AX formation without modifying the levels of ATP in islets. The maximal inhibition of γH2AX formation and loss of ATP in islet cells occur at DPTA/NO concentrations of 400 and 600 μM (Fig. 1B and D) (17, 22). The discordance in ATP levels and inhibition of DDR signaling observed in islet cells treated with 200 μM DPTA/NO compared to those in INS 832/13 cells (Fig. 1) likely reflect the heterogenous nature of endocrine and nonendocrine cells found in islets compared to INS 832/13 cells, which represent a homogenous population of β cells (22). In support of this conclusion, we show that nitric oxide fails to modify ATP levels or attenuate camptothecin-induced DDR signaling in mouse embryonic fibroblasts (MEF) (Fig. 1E and F) and HepG2 cells (Fig. 1G and H). DPTA/NO alone, which fails to stimulate γH2AX formation in insulin-containing cells, induces the formation of γH2AX in non-insulin-containing MEF and HepG2 cells (Fig. 1). These findings associate the cell-type-selective inhibition of DDR signaling by nitric oxide with decreases in the cellular levels of ATP.

FIG 1.

Nitric oxide-dependent inhibition of γH2AX formation in β cells correlates with a decrease in ATP levels. The formation of γH2AX was determined by Western blot analysis (A, C, E, and G) and quantified by densitometry (B, D, F, and H), and ATP levels (B, D, F, and H) were determined following a 2-h incubation of INS 832/13 cells (A and B), MEF (E and F), and HepG2 cells (G and H) with camptothecin (25 μM) or dispersed rat islet cells pretreated for 2 h with DPTA/NO at the indicated concentrations and then treated for 30 min with H₂O₂ (100 μM) (C and D). GAPDH levels were determined to control for protein loading and for normalization of γH2AX formation. Results are representative (A, C, E, and G) or the averages ± standard errors of the means (SEM) (B, D, F, and H) of data from two to four independent experiment and are consistent with our previous findings (17). The inhibition of γH2AX formation by DPTA/NO in response to camptothecin (B) or H₂O₂ (D) (*, P < 0.05) and the DPTA/NO-mediated decrease in ATP (B and D) (#, P < 0.05) achieved statistical significance.

Glycolytic flux in β cells and non-β cells exposed to nitric oxide.

Extracellular flux analysis was performed using the Seahorse XF96 analyzer to investigate the actions of nitric oxide on pathways responsible for ATP generation in β cells and non-β cells. Nitric oxide is an inhibitor of the Krebs cycle enzyme aconitase and complex IV of the electron transport chain (ETC) (27, 28), and in a concentration-dependent manner, it inhibits oxygen consumption (oxygen consumption rate [OCR]) in β cells, HepG2 cells, and MEF (Fig. 2). MEF and HepG2 cells adapt to the inhibition of mitochondrial respiration by increasing glycolytic flux, as indicated by an increase in the extracellular acidification rate (ECAR) (Fig. 2E and F) (29). It is likely that this shift to glycolytic metabolism allows MEF and HepG2 cells to maintain their ATP levels in the presence of nitric oxide (Fig. 1). β cells fail to compensate for the inhibition of mitochondrial respiration with an increase in glycolysis (ECAR) (Fig. 2D), and as a consequence, ATP levels are decreased in response to nitric oxide (Fig. 1). Glycolytic oxidation and mitochondrial oxidation of glucose are tightly coupled in β cells such that >90% of the carbons of glucose are oxidized to CO₂, and this oxidation occurs on the substrate supply (oxidation rates increase in a glucose-dependent manner) (30, 31). These findings, which associate ATP levels with an inhibition in DDR signaling by nitric oxide, suggest that cell type selectivity in this response may reflect an inability of β cells to compensate for impaired mitochondrial respiration with increases in glycolytic metabolism.

FIG 2.

Effects of nitric oxide on oxidative metabolism in β cells and non-β cells. The time- and concentration-dependent effects of DPTA/NO on the rates of oxygen consumption (OCR) (A to C) and extracellular acidification (ECAR) (D to F) were measured in the indicated cell types by extracellular flux analysis. The time of DPTA/NO administration is indicated by arrows. The results are the averages ± SEM of data from three independent experiments.

To further explore this potential hypothesis, the effects of 400 μM DPTA/NO on the OCR and ECAR were compared between INS 832/13 cells, MEF, and HepG2 cells (Fig. 3A and B). This direct comparison highlights that nitric oxide (DPTA/NO) impairs mitochondrial respiration in each of the cell types; however, β cells fail to compensate for this inhibition with an increase in flux through glycolysis, while glycolytic flux (ECAR) is increased in non-β cells (Fig. 3A and B). The lack of glycolytic compensation of β cells is due, in part, to the inability of β cells to regenerate NAD⁺ (Fig. 3C), a required cofactor for the oxidation of glyceraldehyde-3-phosphate to 1,3-bisphosphoglycerate by glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (32). The reduction in NAD⁺ levels correlates with the absence of lactate dehydrogenase (LDH) (33) or detectable levels of LDH activity in INS 832/13 cells (Fig. 3E and F). These findings are consistent with previous reports showing minimal expression of LDH in β cells (23), while non-β cells express and maintain robust LDH activity.

FIG 3.

Effects of nitric oxide on ATP and NAD⁺ levels in β and non-β cells. (A and B) The time-dependent effects of 400 μM DPTA/NO on the OCR (A) and ECAR (B) in MEF, HepG2, and INS 832/13 cells measured by extracellular flux analysis in Fig. 2 were plotted together for direct comparison. The dashed line indicates 100% baseline, and the arrow indicates when DPTA/NO was administered. (C) NAD⁺ levels were determined by HPLC in INS 832/13 cells, MEF, HepG2 cells, and dispersed rat islets treated for 2 h with DPTA/NO (600 μM). (D) ATP levels from experiments in Fig. 1 were plotted alongside NAD⁺ measurements for comparison. (E and F) LDH expression was determined by Western blot analysis (E), and enzymatic activity was measured from INS 832/13, MEF, and HepG2 cells (F). GAPDH was used as a control for protein loading. Results are representative (E) or averages ± SEM (A to D and F) of data from 3 to 5 individual experiments. Differences in NAD⁺ and ATP levels compared to DPTA/NO treatment (C and D) and in LDH activity compared to INS 832/13 cells (F) reached statistical significance (*, P < 0.05).

Mitochondrial inhibitors impair DDR signaling in β cells.

The data presented in Fig. 1 to 3 support a role for the inhibition of mitochondrial oxidation and absence of glycolytic compensation leading to ATP depletion as one mechanism by which nitric oxide attenuates DDR signaling selectively in β cells. To further explore this possibility, the effects of inhibitors of mitochondrial respiration on DDR signaling were examined in β cells and non-β cells. Similar to DPTA/NO, the mitochondrial inhibitors rotenone (complex I), antimycin A (complex III), oligomycin (ATP synthase), and carbonyl cyanide 4-(trifluoromethoxy)phenylhydrazone (FCCP) (uncoupler) attenuate camptothecin-induced KAP1 phosphorylation and γH2AX formation in a concentration-dependent manner (Fig. 4A to F). Like nitric oxide, the inhibition of DDR signaling is associated with an ∼10-fold decrease in ATP levels (Fig. 4G). These findings have been confirmed in primary rat islets, where H₂O₂-induced KAP1 phosphorylation and γH2AX formation are inhibited by rotenone (Fig. 4H and I) under conditions in which there is an ∼75% reduction in islet ATP levels (Fig. 4J). In contrast, inhibitors of mitochondrial respiration at concentrations that attenuate DDR signaling and decrease ATP levels in INS 832/13 cells do not modify DDR activation or the cellular levels of ATP in MEF (Fig. 4K to M). The actions of mitochondrial inhibitors are consistent with the effects of nitric oxide and provide additional evidence that the selective inhibition of DDR signaling in β cells is associated with a decrease in mitochondrial respiration and a lack of glycolytic compensation leading to decreases in ATP levels in β cells.

FIG 4.

Mitochondrial inhibitors prevent DDR signaling in β cells. (A to F) The concentration-dependent effects of rotenone (A), antimycin A (B), oligomycin (C), or FCCP (D) on camptothecin-induced (25 μM) KAP1 phosphorylation and γH2AX formation were determined by Western blot analysis and quantified by densitometry (E and F). The value for GAPDH was determined to control for protein loading, and the effects of DPTA/NO (400 μM) are shown as a positive control for DDR inhibition. (G) The effects of rotenone (1 μM), antimycin A (100 nM), oligomycin (100 nM), and FCCP (10 μM) on INS 832/13 cell ATP levels were determined following a 2-h treatment. (H to J) Dispersed rat islet cells were pretreated for 1 h with rotenone (1 μM), 100 μM H₂O₂ was added, and the cells were cultured for an additional 30 min. Following the treatments, KAP1 phosphorylation and γH2AX formation were determined by Western blot analysis (H) and quantified by densitometry using GAPDH as a loading control (I), and ATP levels were measured by HPLC (J). (K to M) The effects of a 2-h incubation with rotenone (1 μM), antimycin A (100 nM), oligomycin (100 nM), and FCCP (10 μM) on camptothecin (25 μM)-induced γH2AX formation in MEF were assessed by Western blot analysis (K) and quantified by densitometry (L), and ATP levels were quantified under these conditions (M). Results are representative (A to D, H, and K) or the averages ± SEM of data from three (E to G and L) or two (I, J, and M) independent experiments. The inhibition of camptothecin-induced KAP1 phosphorylation and γH2AX formation and the decrease in ATP (E to G) achieved statistical significance (* and #, P < 0.05).

Inhibitors of mitochondrial respiration attenuate DDR signaling in MEF when glycolytic flux is limited.

Since non-β cells maintain ATP levels and DDR signaling when mitochondrial respiration is impaired by increasing glycolytic metabolism (Fig. 2 and 3), the effects of glucose depletion on DDR signaling were examined. MEF were cultured in glucose-free modified Ringer’s buffer containing glutamine as the only carbon source (34). Under these culture conditions, rotenone attenuates camptothecin-induced KAP1 phosphorylation and γH2AX formation (Fig. 5A), while MEF cultured in standard medium maintain DDR signaling in the presence of this inhibitor (Fig. 4K). Consistent with the loss of DDR signaling, the addition of rotenone to MEF cultured in the absence of glucose results in an ∼10-fold decrease in ATP (Fig. 5B), as the cells can no longer compensate for impaired mitochondrial respiration with increased glycolytic flux. Likewise, inhibition of glycolysis using 2-deoxy-d-glucose (2-DG) also attenuates camptothecin-induced DDR activation while stimulating AMPK phosphorylation (a positive control for ATP depletion) in MEF cultured in the absence of glucose (Fig. 5A). These findings show that DDR signaling can be regulated in non-β cells in a manner that is similar to the regulation observed in β cells when glycolytic compensation for impaired mitochondrial respiration is limited.

FIG 5.

Effects of glucose depletion on DDR signaling in MEF. (A) The phosphorylation of AMPK and KAP1 and the formation of γH2AX were determined by Western blot analysis of MEF treated for 1 h with camptothecin (25 μM) in glucose-free modified Ringer’s buffer with and without 2-deoxyglucose (2-DG) (20 mM) and rotenone (10 μM). (B) ATP levels were determined in MEF incubated in glucose-free modified Ringer’s buffer for 1 h in the presence or absence of rotenone. (C to F) The formation of γH2AX was determined by Western blot analysis and quantified by densitometry, and ATP levels were determined by HPLC from MEF (C and D) and INS 832/13 cells (E and F) treated for 1 h with camptothecin in glucose-free modified Ringer’s buffer in the presence or absence of the indicated concentrations of rotenone and glucose. Results are representative (A, C, and E) or averages ± SEM of data from three independent experiments (B, D, and F). The rotenone-mediated decrease in ATP levels (B) (*, P < 0.05) and glucose-stimulated restoration of γH2AX (*, P < 0.05) and ATP (#, P < 0.05) levels (D and F) in MEF and INS 832/13 cells achieved statistical significance.

Since an increase in glycolytic flux is associated with the maintenance of DDR signaling in non-β cells treated with nitric oxide or mitochondrial toxins (Fig. 3B), the effects of increasing concentrations of glucose on ATP levels and DDR signaling in MEF cultured in glucose-free medium and treated with rotenone were explored. In a concentration-dependent manner, glucose restores the formation of γH2AX in rotenone-treated MEF (Fig. 5C and D). At a concentration of 5 mM, glucose is capable of fully restoring γH2AX formation (Fig. 5C and D), while concentrations of glucose above 5 mM did not further augment γH2AX formation in response to camptothecin (not shown). The restoration of γH2AX formation correlates with the accumulation of ATP to levels that are ∼60% of the levels measured in untreated control cells (Fig. 5D). Similar to MEF, camptothecin-induced γH2AX formation is inhibited in INS 832/13 cells cultured in the absence of glucose and exposed to rotenone (Fig. 5E); however, the addition of basal levels of glucose (5 mM) to camptothecin-treated INS 832/13 cells is not sufficient to support γH2AX formation or the restoration of ATP. Increasing the concentration of glucose to 20 mM allows γH2AX to accumulate to levels that are 30% of those observed in camptothecin-treated INS 832/13 cells (Fig. 5E and F). The small increase in γH2AX formation in response to 20 mM glucose correlates with a similar increase in the levels of ATP present in these cells (Fig. 5E and F).

Effects of nitric oxide on DDR signaling in MEF forced to generate ATP via mitochondrial oxidation.

If glycolytic compensation is the mechanism that allows non-β cells to generate sufficient levels of ATP to maintain DDR signaling, then nitric oxide should attenuate DDR signaling if these cells are forced to rely on mitochondria as the primary source of ATP. Galactose is a poor substrate for glycolysis, and cells cultured in galactose shift to mitochondrial oxidation of glutamine as the primary energy-generating pathway (35–37). In fact, mitochondrial respiration (OCR) is enhanced in cells cultured in galactose-containing medium compared to glucose-containing medium, supporting a shift to the mitochondrial oxidation of glutamine under these culture conditions (35–37). While the ECAR is enhanced by DPTA/NO in MEF cultured in glucose, it is not modified by nitric oxide when these cells are cultured in galactose-containing medium (Fig. 6A). Consequently, nitric oxide inhibits camptothecin-induced γH2AX formation in a concentration-dependent manner (Fig. 6B and C) that correlates with the depletion of ATP in MEF cultured in galactose-containing medium (Fig. 6D). When cells are cultured in glucose, nitric oxide does not modify camptothecin-induced γH2AX, nor does it decrease cellular levels of ATP in MEF (Fig. 6B and D). In fact, nitric oxide alone stimulates γH2AX formation in MEF cultured in glucose, while nitric oxide fails to stimulate γH2AX formation in galactose-cultured MEF (Fig. 6B).

FIG 6.

The role of metabolic flexibility in the regulation of DDR signaling by nitric oxide. (A) MEF, cultured in 10 mM glucose or 10 mM galactose, were treated with or without 400 μM DPTA/NO for 2 h, and the ECAR was determined by extracellular flux analysis. ECAR values obtained in INS 832/13 cells from experiments in Fig. 3 are shown for comparison. (B to D) The concentration-dependent effects of DPTA/NO on camptothecin (25 μM)-induced γH2AX formation were determined by Western blot analysis (B) and quantified by densitometry (C), and ATP levels were determined on MEF cultured in glucose- or galactose-containing medium (D). (E and F) The effects of a 2-h incubation with rotenone (R), antimycin A (A), oligomycin (O), or FCCP (F) on camptothecin-induced γH2AX formation and KAP1 phosphorylation were determined by Western blot analysis (E) and quantified by densitometry (F) in MEF cultured in glucose- or galactose-containing medium. (G and H) The time-dependent effects of 400 μM DPTA/NO on the phosphorylation of KAP1 and eIF2α and the formation of γH2AX (G and H) and ATP levels (H) were determined in MEF cultured in glucose- or galactose-containing medium. Results are representative (B, E, and G) or averages ± SEM of data from three to seven independent experiments. Statistically significant differences are indicated (*, P < 0.05).

Results presented in Fig. 6A to D indicate that forcing MEF to use mitochondria as the primary source for ATP generation (when cultured in galactose and glutamine) transforms the phenotype of these cells to one that is more consistent with β cells (Fig. 6). In further support of this conclusion, we show that the mitochondrial respiratory chain inhibitors rotenone, antimycin A, oligomycin, and FCCP suppress camptothecin-induced KAP1 phosphorylation and γH2AX formation in MEF cultured in galactose-containing medium (Fig. 6E and F). These inhibitors of mitochondrial respiration do not modify DDR signaling in MEF grown in glucose-containing medium or under conditions in which these cells maintain ATP levels (Fig. 6E and F).

The effects of changes in metabolism on the ability of nitric oxide to modify signaling cascades is not limited to the DDR. We have shown that under conditions in which nitric oxide inhibits DDR signaling in β cells, the unfolded protein response (UPR) is activated, as evidenced by α subunit of eukaryotic initiation factor 2 (eIF2α) phosphorylation (22). We now show that nitric oxide induces eIF2α phosphorylation in MEF cultured in galactose-containing medium (Fig. 6G) or under conditions in which nitric oxide also inhibits DDR signaling (Fig. 6B). When DDR signaling is intact in glucose-cultured MEF (KAP1 phosphorylation and γH2AX formation), DPTA/NO fails to stimulate eIF2α phosphorylation (Fig. 6G). Furthermore, the inhibitory actions of nitric oxide on DDR signaling and stimulatory actions on eIF2α phosphorylation in galactose-cultured MEF correlate with the loss of cellular ATP, while ATP levels are not changed by nitric oxide in MEF cultured in glucose (Fig. 6H). These findings indicate that cells that can switch between glycolysis and mitochondrial respiration to sustain energy demands in the presence of nitric oxide have the capacity to activate DDR signaling, while cells that lack this metabolic flexibility (β cells or MEF maintained on galactose) cannot sustain DDR signaling when mitochondrial respiration is impaired.

Nitric oxide protects galactose-cultured MEF from DNA damage-induced cell death.

Since irreversible DNA damage can result in DDR-mediated apoptosis (19), and nitric oxide attenuates DDR-dependent β-cell apoptosis (17), the effects of nitric oxide on DDR-induced apoptosis were examined in MEF cultured in galactose. In glucose-cultured MEF, DPTA/NO and camptothecin stimulate caspase-3 activation (cleavage), and in combination, the actions of each stimulus are additive (Fig. 7A and B). In contrast, MEF cultured in galactose are resistant to DPTA/NO-induced caspase-3 activation, and nitric oxide attenuates camptothecin-induced caspase-3 cleavage (Fig. 7A and B). Furthermore, MEF cultured in galactose are resistant to DPTA/NO-induced cell death, while DPTA/NO reduces the viability of MEF cultured in glucose in a concentration-related manner, with 15% killing at 200 μM and 25% death at 400 μM nitric oxide (Fig. 7C). Much like caspase activation (Fig. 7A), camptothecin stimulates the apoptosis of MEF to similar levels whether cultured in glucose or galactose (Fig. 7D). Importantly, DDR inhibition by nitric oxide attenuates camptothecin-induced apoptosis only in galactose-cultured MEF (Fig. 7D), consistent with the protective effects of nitric oxide on camptothecin-induced apoptosis of INS 832/13 cells (Fig. 7E) (17). Nitric oxide does not prevent camptothecin-induced cell death in MEF cultured in medium containing glucose (Fig. 7D). These findings suggest that metabolic flexibility is essential for maintaining DDR signaling and DDR-dependent apoptosis under conditions where mitochondrial respiration is impaired. In cells that can adapt to impaired mitochondrial respiration with increased glycolytic flux, DDR signaling and DDR-dependent apoptosis are not modified by nitric oxide; however, in cells that lack the ability to compensate for impaired mitochondrial respiration with increased glycolysis, such as β cells, nitric oxide attenuates DDR signaling and DDR-dependent apoptosis.

FIG 7.

The role of metabolic flexibility in the regulation of DNA damage-induced cell death. (A to C) The effects of DPTA/NO (300 μM) on camptothecin (25 μM)-induced cleavage of caspase-3 were determined by Western blot analysis (A) and quantified by densitometry (B), and SYTOX green nucleic acid stain was used to measure cell death (C) in MEF cultured in glucose- or galactose-containing medium (A and B) with the indicated concentration of this nitric oxide donor (C). (D) The concentration-dependent effects of DPTA/NO on camptothecin-induced MEF cell death were determined following a 24-h incubation in glucose- or galactose-containing medium. (E) The effects of DPTA/NO (300 μM) on camptothecin-induced INS 832/13 cell death were determined following a 24-h incubation. Results are representative (A) or the averages ± SEM (B to E) of data from three or four independent experiments. Statistically significant differences are indicated (*, P < 0.05).

DISCUSSION

A central function of the DDR is to coordinate the cellular responses to DNA damage (19, 20). We have shown that nitric oxide acts as both a positive and a negative regulator of DDR signaling in β cells (15, 17, 18). Nitric oxide, a free radical with genotoxic properties, causes DNA damage leading to DDR activation in multiple cell types, including β cells (18, 22, 38–41). However, under conditions where nitric oxide is present at low-micromolar levels, such as during exposure to proinflammatory cytokines, DDR signaling in β cells is suppressed (17, 22). Nitric oxide inhibits the phosphorylation of p53 and KAP1 and the formation of γH2AX in response to DNA-damaging agents such as hydrogen peroxide and camptothecin, despite the presence of DNA damage (17). While inhibiting the DDR, nitric oxide also activates stress responses, including mitogen-activated protein kinase (MAPK), AMPK, the UPR, and heat shock (42). The inhibition of DDR signaling by micromolar levels of nitric oxide is restricted to insulin-producing β cells and results in the attenuation of DDR-dependent apoptosis (17).

In many physiological settings, nitric oxide plays both protective and damaging roles that depend on its interactions with other radicals, the amount produced, and the function of the tissue producing the radical. We and others have shown that cytokine-induced inhibition of oxidative metabolism, insulin secretion, and the loss of β-cell viability can be prevented by NOS inhibition (4–6, 43), and islets isolated from iNOS knockout mice are resistant to cytokine-mediated damage (44). Nitric oxide also stimulates the expression of genes that participate in functional recovery from cytokine-mediated damage, including the recovery of metabolic function and the repair of damaged DNA (16, 45, 46). These protective and damaging actions are consistent with the ability of nitric oxide to activate and inhibit DDR signaling selectively in β cells (17, 18, 22, 38–41).

In this study, the mechanisms responsible for the cell type specificity of DDR inhibition by nitric oxide were examined. We have previously shown that de novo gene transcription, new protein synthesis, and UPR activation do not participate in the regulation of DDR signaling by nitric oxide in β cells (22). Also, phosphatases known to negatively regulate DDR signaling, such as PP1 and PP2A or the inhibitor of protein phosphatase 1, a phosphatase inhibitor selectively expressed in β cells (22, 47), do not participate in DDR regulation by nitric oxide. Since the inhibition in DDR signaling by nitric oxide is selective for β cells, we examined the role of pathways unique to β cells. The regulation of metabolism in β cells is unique in that glycolysis and mitochondrial oxidative metabolism are coupled, such that >90% of the carbons of glucose are oxidized to CO₂ (23, 30). The control of this metabolism is based on the cell-type-selective expression of glucokinase and GLUT2 in β cells and the metabolic coupling of glycolysis and mitochondrial oxidation of glucose (25). This allows the β cell to use the oxidative metabolism of glucose as a sensor that allows for exquisite control of insulin secretion in a manner that is proportional to changes in extracellular glucose levels (25, 26, 30). In most cell types, glycolysis and mitochondrial respiration are uncoupled, allowing these cell types to rely on glycolysis to maintain energy stores (26). In this report, we provide evidence that the selective inhibition of DDR signaling by nitric oxide in β cells is associated with limited metabolic flexibility to shift to glycolytic metabolism for energy needs when mitochondrial respiration is impaired (Fig. 8).

FIG 8.

Metabolic flexibility and regulation of DDR signaling by nitric oxide. Evidence presented in this study supports the loss of cellular levels of ATP as the mechanism responsible for the inhibition of DDR signaling selectively in β cells. Under conditions of impaired mitochondrial oxidation due to the actions of nitric oxide, β cells lack the metabolic flexibility to maintain ATP levels through glycolysis. Non-β cells sustain ATP production through glycolysis, allowing these cells to maintain DDR signaling under conditions in which nitric oxide inhibits mitochondrial oxidation.

Nitric oxide is an effective inhibitor of mitochondrial respiration (28) through Fe-S cluster disruption and/or S-nitrosation of complex I (48), reversible occupation of the oxygen binding site in complex IV (49, 50), and destruction of the 4Fe-4S cluster of aconitase (27). Most cell types compensate for an impairment in mitochondrial respiration with an increase in glycolysis. Indeed, nitric oxide inhibits the consumption of oxygen by HepG2 cells and MEF, and these cells compensate by increasing glycolysis. This compensation allows HepG2 cells and MEF to maintain ATP levels in the presence of inhibitors of mitochondrial oxidative metabolism (Fig. 2). β cells lack the ability to adapt to impaired mitochondrial oxidation with increased glycolysis, as nitric oxide decreases NAD⁺ levels in addition to ATP, and NAD⁺ is required to maintain glycolytic flux (GAPDH). LDH is a source of NAD⁺ when oxygen is limiting; however, β cells do not express this enzyme (Fig. 3) (23), consistent with a loss of NAD⁺ and ATP in response to nitric oxide (Fig. 4) (32).

Similar to the actions of nitric oxide, inhibitors of mitochondrial respiration decrease ATP levels and attenuate DDR signaling selectively in β cells (Fig. 4). Since it is known that β cells fail to maintain glycolytic flux in the presence of ETC inhibitors (23), these findings suggest that DDR signaling is impaired in cell types or under conditions in which cells cannot compensate for an impaired mitochondrial oxidative capacity with an increase in glycolytic flux. Consistent with this hypothesis, nitric oxide and mitochondrial inhibitors attenuate DDR signaling in MEF cultured in the absence of glucose as well as in MEF cultured in galactose. Galactose is not effectively metabolized via glycolysis, and when cells are cultured in galactose (absence of glucose), they generate ATP via mitochondrial glutamine metabolism (35–37). In addition to inhibiting DDR signaling, nitric oxide also attenuates DDR-mediated apoptosis selectively in β cells (17). We now show that nitric oxide is capable of inhibiting DDR-mediated apoptosis of MEF cultured in galactose (Fig. 7) or under conditions in which ATP is primarily derived from mitochondrial metabolism. Under conditions in which DDR-induced apoptosis is attenuated, nitric oxide activates the UPR in both β cells (22) and MEF cultured in galactose (Fig. 6). Nitric oxide does not inhibit DDR signaling or prevent the apoptosis of MEF cultured in glucose-containing medium (Fig. 7). In fact, nitric oxide alone is capable of activating DDR signaling (Fig. 6) and decreasing MEF viability (Fig. 7). Much like nitric oxide, inhibitors of mitochondrial oxidative metabolism (rotenone) also attenuate DDR signaling in MEF cultured in the absence of glucose or in medium containing galactose and glutamine as the primary carbon sources (Fig. 6). Under these conditions, there is a decrease in cellular ATP levels; however, the levels of ATP are above the K_m for ATM, as evidenced by the activation of kinases (AMPK, pancreatic endoplasmic reticulum kinase [PERK], and extracellular signal-regulated kinase [ERK]) with a similar K_m for ATP (51).

These findings suggest that metabolic flexibility is essential in regulating DDR signaling (Fig. 8). In cells where glycolysis and mitochondrial oxidation are coupled, or under conditions in which mitochondrial respiration is required for ATP generation, nitric oxide and inhibitors of mitochondrial respiration attenuate DDR signaling and DDR-dependent apoptosis. In cell types with metabolic flexibility, DDR signaling is not modified by inhibitors of mitochondrial oxidative metabolism such as nitric oxide or rotenone. This form of DDR regulation appears to be a protective response that prevents β cells from undergoing premature apoptosis in response to DNA damage that may occur during islet inflammation. It also represents an intrinsic protective mechanism that couples cell-type-specific regulation of intermediary glucose metabolism with the pathways that control cell fate in response to stressors. These findings may also provide mechanistic insights into the potential role of intermediary metabolism as a contributor to the efficacy of DNA-damaging agents in the killing of cancer cells.

MATERIALS AND METHODS

Materials and animals.

Male Sprague-Dawley rats were purchased from Harlan (Indianapolis, IN). Cell lines used (and suppliers) include INS 832/13 cells (Chris Newgard, Duke University, Durham, NC), mouse embryonic fibroblasts (MEF; Fumihiko Urano, Washington University, St. Louis, MO), and HepG2 cells (American Type Culture Collection). Minimum essential medium (MEM), Connaught Medical Research Laboratories (CMRL) 1066 medium, d-glucose, and β-mercaptoethanol were purchased from Thermo Fisher Scientific (Waltham, MA). RPMI 1640 medium, Dulbecco’s modified Eagle medium (DMEM), trypsin (0.05% in 0.53 mM EDTA), l-glutamine, sodium pyruvate, HEPES, penicillin, and streptomycin were purchased from Corning (Corning, NY). Fetal bovine serum was purchased from GE Healthcare Life Sciences (Marlborough, MA). The nitric oxide donor (Z)-1-[N-(3-aminopropyl)-N-(3-ammoniopropylamino)]diazen-1-ium-1,2-diolate (DPTA/NO) and oligomycin were purchased from Cayman Chemical (Ann Arbor, MI). H₂O₂, camptothecin, 2-deoxy-d-glucose (2-DG), antimycin A, rotenone, carbonyl cyanide 4-(trifluoromethoxy)phenylhydrazone (FCCP), and d-galactose were purchased from Sigma-Aldrich (St. Louis, MO). The antibodies used (and their sources) were as follows: rabbit anti-phospho-KAP1 (Ser-824) and rabbit anti-LDH (Abcam, Cambridge, MA); mouse anti-phospho-H2AX (Ser-139; γH2AX) (EMD Millipore, Billerica, MA); mouse anti-GAPDH (glyceraldehyde-3-phosphate dehydrogenase) (Thermo Fisher Scientific); rabbit anti-phospho-AMPK (Thr-172), rabbit anti-cleaved caspase-3, and rabbit anti-phospho-eIF2α (Ser-51) (Cell Signaling Technology, Beverly, MA); and horseradish peroxidase (HRP)-conjugated donkey anti-rabbit and HRP-conjugated donkey anti-mouse antibodies (Jackson ImmunoResearch Laboratories, Inc., West Grove, PA).

Rodent islet isolation and cell culture.

Islets were isolated from Sprague-Dawley rats and cultured as previously described (52). INS 832/13, MEF, and HepG2 cells were cultured as previously described (17, 18, 53). All animal care and experimental procedures involving rodents were approved by the Institutional Animal Care and Use Committees at the Medical College of Wisconsin (A3102-01).

Western blot analysis.

Cells and islets were lysed directly in Laemmli buffer, and Western blot analysis was conducted as previously described (54). The following dilutions of primary and secondary antibodies were used: 1:1,000 for cleaved caspase-3, phospho-eIF2α, and phospho-AMPK; 1:5,000 for γH2AX; 1:2,000 for phospho-KAP1; 1:20,000 for GAPDH; 1:2,000 for LDH; and 1:20,000 for donkey anti-mouse and donkey anti-rabbit antibodies conjugated to horseradish peroxidase. Bands were detected using chemiluminescence (55).

Nucleotide measurements.

Cellular levels of ATP and NAD⁺ were quantified using high-performance liquid chromatography (HPLC) as previously described (56, 57). Nucleotides were extracted by perchloric acid precipitation (58), and solvent A (75 μl; 0.1 M potassium phosphate and 4 mM tetrabutylammonium bisulfate [pH 6.0], diluted 64:36 [vol/vol] in water) was added to supernatants. Precipitated protein was isolated by centrifugation and solubilized in 0.5 N NaOH. The protein concentration was determined using the bicinchoninic acid (BCA) protein assay (Thermo Fisher Scientific). HPLC analysis of nucleotides was performed on a Supelcosil LC-18-T column (3 μm; 150- by 4.6-mm internal diameter) according to previously described methods (59). ATP and NAD⁺ peaks were measured for each sample and expressed in nanomoles per milligram of protein.

Cellular bioenergetics.

INS 832/13 cells (20,000 cells/well), MEF (10,000 cells/well), and HepG2 cells (10,000 cells/well) were plated in a Seahorse XF96 cell culture microplate, and extracellular flux was assessed using a Seahorse XF96 analyzer. Unless indicated otherwise, experiments were conducted with DMEM containing 5.5 mM glucose, 2 mM pyruvate, and 1 mM glutamine. Changes in the oxygen consumption rate (OCR) and extracellular acidification rate (ECAR) in response to nitric oxide are expressed as a percentage of the baseline for each cell type.

Lactate dehydrogenase activity assay.

LDH activity was measured as previously described (23), with slight modifications. Cells were collected in ice-cold phosphate-buffered saline (PBS) and lysed via sonication. The lysates were centrifuged at 16,000 × g at 4°C for 5 min and added to assay buffer containing 100 mM HEPES (pH 7.4), 0.05% (wt/vol) bovine serum albumin (BSA), 4 mM pyruvate, and 40 μM NADH, and the activity was measured by monitoring the NADH absorbance at 340 nm using a BioTek Synergy MX plate reader. Activity was normalized to total protein, measured by the BCA protein assay.

ATP depletion.

ATP depletion was performed as previously described (34) by replacing cell culture medium with modified Ringer’s buffer (115 mM NaCl, 5 mM K₂HPO₄, 25 mM NaHCO₃, 2 mM MgSO₄, 1 mM CaCl₂, 2 mM glutamine [pH 7.4]) in the presence of 2-DG or rotenone at the concentrations indicated in the figure legends.

Cell death assay.

Cell death was measured by SYTOX green uptake (Thermo Fisher Scientific) as previously described (17). Following treatment, cells were incubated with SYTOX green at a final concentration of 5 μM for 30 min at 37°C. Fluorescence was then determined at an excitation/emission wavelength of 504/523 nm. Percent cell death is based on conditions of 100% cell death as measured from digitonin (120 μM)-permeabilized cells.

Statistical analysis.

Statistical analysis was performed using one-way analysis of variance with a Tukey-Kramer post hoc test, with significance at a P value of <0.05, as indicated in the figure legends.