Induction of Fatty Acid Oxidation Underlies DNA Damage‐Induced Cell Death and Ameliorates Obesity‐Driven Chemoresistance

Abstract The DNA damage response is essential for preserving genome integrity and eliminating damaged cells. Although cellular metabolism plays a central role in cell fate decision between proliferation, survival, or death, the metabolic response to DNA damage remains largely obscure. Here, this work shows that DNA damage induces fatty acid oxidation (FAO), which is required for DNA damage‐induced cell death. Mechanistically, FAO induction increases cellular acetyl‐CoA levels and promotes N‐alpha‐acetylation of caspase‐2, leading to cell death. Whereas chemotherapy increases FAO related genes through peroxisome proliferator‐activated receptor α (PPARα), accelerated hypoxia‐inducible factor‐1α stabilization by tumor cells in obese mice impedes the upregulation of FAO, which contributes to its chemoresistance. Finally, this work finds that improving FAO by PPARα activation ameliorates obesity‐driven chemoresistance and enhances the outcomes of chemotherapy in obese mice. These findings reveal the shift toward FAO induction is an important metabolic response to DNA damage and may provide effective therapeutic strategies for cancer patients with obesity.


Introduction
Obesity, a physical condition defined as abnormal or excessive accumulation of fat, is increasing worldwide. [1]4a] It is noteworthy that obese patients recapitulate a state of persistent low-grade inflammation characterized by increased cytokines and chemokines that induce alterations in tumor microenvironment. [8]4a,8b] However, it remains poorly understood whether overweight or obesity adversely affects therapeutic outcomes directly through intrinsic changes of cancer cells.
When cells are exposed to genotoxic stresses, they respond by initiating tightly coordinated signaling responses that can promote survival by inducing cell cycle arrest and DNA repair or trigger cell death to limit the accumulation of damaged cells. [9]iven the indispensability of cellular metabolism for survival and growth, cells exposed to DNA damage need to make fundamental metabolic decisions of whether to upregulate biosynthesis for promoting DNA repair or not, and whether to restrict catabolic pathways for limiting proliferation or not.Surprisingly, the metabolic response to DNA damage has been largely unknown.Recent studies have shown that there is a link between metabolism and DNA damage response, by repression of glutamine metabolism to promote proper DNA repair and cell cycle arrest, and to suppress tumorigenesis, [10] implying that there is a metabolic area of cellular stress response with significant therapeutic potential.
Here, we investigated how obesity modulates metabolic stress response of cancer cells to promote cell survival and chemoresistance.Using a murine model of obesity, we performed transcriptomic analyses of metabolic pathways in engrafted cancer cells after chemotherapy.We found that cancer cells in lean mice, but not in obese mice, dynamically respond to chemotherapy by upregulating fatty acid oxidation (FAO) pathway.The FAO induction promotes N-alpha-acetylation of caspase-2 (CASP2), leading to apoptotic cell death.Importantly, we show that pharmacological induction of FAO alleviates obesity-driven drug resistance and enhances therapeutic outcomes after chemotherapy.These findings may have important implications for future therapeutic approaches treating cancer patients with overweight or obesity.

Obesity Impairs FAO Induction in Tumors after Chemotherapy
In order to investigate how obesity induces chemoresistance of tumor cells, we performed allograft tumor formation assays in wild-type (WT) and genetic leptin-deficient ob/ob mice using syngeneic B16F10 melanoma cells.When the tumor volume reached ≈100 mm 3 , mice were administrated with etoposide (ETS), which is one of the most common chemotherapeutic agents and has been shown to induce cell death across multiple cancers. [11]As observed previously, whereas ETS treatment induced cell death in tumors from lean control mice, tumors from obese mice exhibited profound resistance to ETS (Figure 1a).
To explore mechanism(s) underlying the obesity-induced drug resistance, we performed a comparative transcriptomic analysis of tumor cells from lean and obese mice after chemotherapy.We analyzed the following four groups: lean tumor, lean tu-mor after ETS treatment, obese tumor, obese tumor after ETS treatment.As KEGG pathway enrichment analysis, performed on differentially expressed genes (DEGs) in lean tumors before and after chemotherapy (2418 genes; |Fc| ≥ 2 & p-value cutoff < 0.05), suggested that metabolic pathways are the most significantly changed in lean tumors after chemotherapy (Figure 1b).Subsequent pathway mapping revealed that 16 KEGG metabolic signatures significantly changed in lean tumors after chemotherapy (Figure 1c).Importantly, tumors from obese mice also underwent significant changes in these pathways after chemotherapy except 3 pathways including inositol phosphate metabolism, glycine, serine and threonine metabolism, and fatty acid degradation (Figure 1c).Among these differentially regulated pathways, we reasoned that fatty acid metabolism could be most relevant to obesity given its connection to abnormalities of lipid metabolism.
To confirm that obesity abrogates the induction of fatty acid degradation after chemotherapy, we examined gene score enrichment analysis (GSEA) of DEGs in tumors from lean and obese mice pre-and post-chemotherapy.The transcripts involved in fatty acid metabolism were particularly increased in lean tumors after chemotherapy, whereas these changes were not significant in obese tumors (Figure S1a, Supporting Information).Additionally, we observed that the expression of genes involved in FAO pathway was more significantly changed in lean tumors than in obese tumors after chemotherapy (Figure 1d; Figure S1b, Supporting Information).Furthermore, consistent with our transcriptomic analysis, quantitative RT-PCR results confirmed that ETS treatment significantly increased the expression of FAOrelated genes in lean tumors, whereas the induction of these genes was markedly attenuated in obese tumors (Figure 1e,f).Taken together, our findings demonstrate that obesity impairs the induction of FAO gene signatures in tumors after chemotherapy.

Induction of FAO is Required for DNA Damage-Induced Cell Death
Chemotherapies target rapidly proliferating cancer cells by directly or indirectly inducing DNA damage. [12]Because the expression of FAO related genes was highly induced after ETS treatment in lean tumors but not in obese tumors, we hypothesized that FAO could serve as an important regulator of DNA damageinduced cell death after chemotherapy.First, the relevance of FAO induction was further tested using another chemotherapeutic agent, camptothecin (CPT).In line with our in vivo results, genes involved in fatty acid degradation were significantly increased upon ETS or CPT treatment.On the contrary, genes involved in fatty acid synthesis were decreased under these conditions, suggesting that the induction of FAO related genes may be a specific metabolic response to chemotherapy and not reflective of a nonspecific metabolic crisis (Figure 2a).When we directly measured FAO rates, the oxidation of palmitate, a saturated long-chain fatty acid, was significantly increased after ETS or CPT treatment (Figure 2b).
Next to assess whether FAO affects cell death after DNA damage, we inhibited FAO using etomoxir (ETO), which can inhibit mitochondrial import of fatty acids by inhibiting carnitine palmitoyltransferase 1 family proteins (CPT1A, B and C).Notably, inhibition of FAO reduced apoptotic cell death after ETS treat-  ).Statistical analysis was performed using two-way ANOVA with Dunnett's multiple comparisons test.b) FAO was measured in HepG2 cells treated with CPT or ETS (n = 3).Statistical analysis was performed using one-way ANOVA with Dunnett's multiple comparisons test.c) Cell death of B16F10 cells treated with ETS (25 μm, 40 h), ETO (50 or 200 μm, 40 h) or both (n = 3).Cell death was measured by propidium iodide exclusion assay.d) CASP3/7 GLO assay of B16F10 cells treated with ETS (25 μm, 24 h), ETO (50 μm, 24 h) or both (n = 4).e) Cell death of B16F10 cells transfected with nontargeting siRNA (siControl) or with siRNA against CPT1A (siCPT1A).Cells were treated with or without ETO (50 μm, 40 h) in the presence of ETS (25 μm, 40 h) (n = 3).Cell death was measured by propidium iodide exclusion assay.f) Relative Cd36 mRNA level in B16F10 cells treated with ETS (25 μm, 24 h); n = 4. Statistical analysis was based on two-tailed Student's t-test.g) FAO-dependent ΔOCR was measured in B16F10 cells pre-treated with or without palmitate (100 μm, 6 h); n = 5.Statistical analysis was based on two-tailed Student's t-test.h) Cell death of B16F10 cells treated with ETS (25 μm, 40 h) in the presence of palmitate (100 μm, 40 h); n = 3. Statistical analysis was performed using one-way ANOVA with Tukey's multiple comparisons test (c-e and h).All error bars ±SD.*p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001.ment (Figure 2c,d).To further confirm the role of FAO in cell death, we suppressed CPT1A expression by using short interfering RNA (siRNA) and examined the effect on cell survival followed by ETS treatment.Knockdown of CPT1A significantly inhibited DNA damage-induced cell death, and ETO treatment did not have addictive effects on cell death (Figure 2e; Figure S2a,b, Supporting Information).This effect was not specific to B16F10 melanoma cells.We observed similar results in immortalized embryonic fibroblasts (MEFs) (Figure S2c-f, Supporting Information).Additionally, comparable results were obtained in cells treated with other DNA damaging agents, including cisplatin or irradiation (Figure S2g,h, Supporting Information).These results demonstrate that blocking FAO protects cells from DNA damage-induced cell death.
Next, we tested whether increasing FAO potentiates cell death in response to DNA damage.First, we found that the expression of CD36, fatty acid translocase/receptor, was markedly induced following DNA damage (Figure 2f), in line with the induction of FAO.After palmitate treatment, cells exhibited an increased oxygen consumption rate (OCR), indicating that cells were oxidizing more fatty acids (Figure 2g; Figure S2i, Supporting Information).Interestingly, cells treated with palmitate exhibited significantly elevated levels of cell death following ETS treatment (Figure 2h).We obtained comparable results in immortalized MEFs treated with palmitate or oleic acid (Figure S2j,k, Supporting Information).Finally, to test whether the relationship between FAO and cell death is unique to DNA damage or a generalizable feature of other cell death triggers, we treated cells with tumor necrosis factor  (TNF) and cycloheximide, which induces apoptotic cell death through the engagement of death receptors.Unlike DNA damage-induce cell death, cell death induced by activation of death receptors was not reduced by FAO inhibition (Figure S2l, Supporting Information).Taken together, these data illustrate that genotoxic stress induces FAO, which is required for DNA damage-induced cell death.

DNA Damage Induces FAO through PPAR𝜶
As our results show that DNA damaging drugs induce FAO, we sought to identify the molecular mechanism.Peroxisome proliferator-activated receptor  (PPAR) is a critical regulator of lipid metabolism and promotes transcription of genes crucial for FAO, [13] thus suggesting that DNA damage might potentially induce this metabolic response via PPAR.To test this concept, we first assessed the expression of PPAR in response to DNA damage.We observed that ETS treatment resulted in a robust increase in PPAR at mRNA level (Figure 3a), as well as the protein level (Figure 3b) in B16F10 cells and multiple cell lines (Figure S3a-d, Supporting Information).Consistent with these results, the transcriptional activity of PPAR was significantly increased following DNA damage (Figure 3c).
Next, to evaluate the role of PPAR on FAO induction after DNA damage, we treated cells with either GW6471, an antagonist of PPAR, or WY14643, a synthetic agonist of PPAR.Importantly, we observed that PPAR inhibition by GW6471 abrogated the induction of FAO gene signatures upon DNA damage (Figure 3d; Figure S3e, Supporting Information).We obtained similar results in cells in which PPAR expression was reduced by siRNAs against PPAR (Figure S3f,g, Supporting Information).As the PPAR family comprises three subtypes, such as PPAR, PPAR, and PPAR, [14] we investigated the involvement of PPAR or PPAR in the induction of FAO-related genes upon DNA damage.While PPAR knockdown had no effect on the expression of these genes, PPAR knockdown partially reduced the induction of some genes (Figure S3f,h, Supporting Information).However, we observed that PPAR expression remained unaffected after ETS treatment (Figure S3i, Supporting Information), implying that PPAR may play a major role in the induction of FAO upon DNA damage.Consistent with this observation, inhibition of PPAR suppressed the induction of FAO upon DNA damage (Figure 3e).Moreover, to further confirm the in vivo relevance of PPAR, we assessed whether PPAR is induced in tumors after exposure to chemotherapy.Importantly, PPAR was significantly induced in lean tumors after ETS treatment, whereas not in obese tumors (Figure 3f, g; Figure S3j, Supporting Information).These findings prompted us to further investigate whether enhancing PPAR activity rescues the defective FAO induction in obese tumors.We found that administration of WY14643 was able to rescue the impaired induction of FAO related genes in obese tumors after chemotherapy, whereas we did not detect additional effect in lean tumors (Figure 3h).
Given the function of PPAR in FAO regulation, we speculated that the PPAR-mediated transcriptional activation of FAOrelated genes promotes cell death after chemotherapy.To test this hypothesis, we examined the effect of PPAR inhibition on cell survival after DNA damage.Notably, GW6471 treatment markedly reduced DNA damage-induced apoptotic cell death, whereas WY14643 had minimal effect on cell death (Fig 3i; Figure S4a-d, Supporting Information).Additionally, these effects were diminished in CPT1A knockdown cells (Figure S4e, Supporting Information).Similarly, knockdown of PPAR resulted in decreased cell death following ETS treatment compared to control cells (Figure S4f,g, Supporting Information).To further confirm the importance of this pathway in DNA damageinduced cell death, we performed allograft assays with PPAR knockdown melanoma cells to assess changes in drug-sensitivity.Consistent with our in vitro results, PPAR knockdown led tumors to be more resistant to ETS treatment (Figure 3j).Together, these data support the idea that DNA damage induces FAO via PPAR, which promotes cell death.

Obesity-Induced Hypoxia Suppresses FAO Induction after Chemotherapy
Because obese tumors are more hypoxic than lean tumors, [15] and hypoxia-inducible factor-1 (HIF1) suppresses FAO by inhibiting PPAR, [16] we hypothesized that hypoxic condition occurring in obese tumors would contribute to the defect in FAO induction after chemotherapy.To test this idea, we first investigated whether obesity augments hypoxia in tumors.As has previously shown, [15a] histological analysis showed that tumors from obese mice exhibited elevated levels of HIF1 relative to tumors from lean mice (Figure 4a; Figure S5a, Supporting Information).To test whether hypoxia directly represses FAO induction after DNA damage, we examined the levels of FAO related genes under hypoxic conditions.ETS treatment clearly and repro- ducibly increased the expression of PPAR and its target genes in normoxia (Figure 4b-d; Figure S5b, Supporting Information).However, the induction of these genes was markedly blunted by hypoxia (Figure 4b-d; Figure S5b, Supporting Information).We also obtained similar results in cells treated with dimethyloxalglycine (DMOG), which stabilizes HIF1 by inhibiting prolyl hydroxylase (Figure S5c, Supporting Information).Next, to further investigate whether HIF1 directly represses FAO induction after DNA damage, we reduced HIF1 expression by siRNA targeting HIF1 and examined the expression of FAOrelated genes under hypoxic conditions.Notably, knockdown of HIF1 significantly restored the impaired induction of PPAR and FAO-related genes after ETS treatment (Figure 4e,f; Figure S5d, Supporting Information).In conclusion, these data suggest that obesity-mediated hypoxia opposed the FAO induction after chemotherapy.

FAO Regulates DNA Damage-Induced Cell Death by Promoting N-Alpha-Acetylation of Caspase-2
We asked how FAO regulates cell death after chemotherapy.First, to examine the involvement of p53 in FAO-mediated cell death, we first reduced p53 expression using siRNA against p53.Notably, ETO suppressed cell death in p53 knockdown cells after ETS treatment (Figure 5a; Figure S6a, Supporting Information), implying that FAO might regulate DNA damage-induced cell death in a p53-independent manner.Consistent with this observation, we observed similar results in p53 null PC3 human prostate cancer cells (Figure S6b, Supporting Information).Previous studies have shown that caspase-2 (CASP2) is required for apoptosis in p53-deficient cells. [17]Thus, to investigate the contribution of CASP2 to the FAO-mediated cell death, we examined CASP2 activity under these conditions.Interestingly, FAO inhibition markedly reduced levels of its cleaved forms after DNA damage (Figure 5b; Figure S6c,d, Supporting Information).Consistent with this observation, ETO was not able to further reduce cell death in CASP2 knockdown cells (Figure 5c; Figure S6e,f, Supporting Information), indicating that FAO might induce cell death by activating CASP2 after chemotherapy.
Previous studies have found that N-alpha-acetylation of CASP2 is required for its activation, which is dependent on cellular levels of acetyl-CoA. [18]Because FAO is an important source of cellular acetyl-CoA pools, [19] we hypothesized that FAO regulates DNA damage-induced cell death by promoting CASP2 activity through acetyl-CoA.To test this idea, we first examined whether FAO induction induces cell death via acetyl-CoA upon DNA dam-age.Importantly, we found that ETS treatment elevated intracellular acetyl-CoA levels (Figure 5d; Figure S6g, Supporting Information).The induction of acetyl-CoA levels required FAO, as ETO treatment blocked this (Figure 5d; Figure S6g, Supporting Information).As citrate can be converted to acetyl-CoA by ATP citrate lyase, we suspected that the decreased cell death by ETO could be restored by citrate.As expected, addition of citrate increased cellular acetyl-CoA levels (Figure 5d; Figure S6g, Supporting Information).Notably, citrate treatment rescued the reduced DNA damage-induced cell death by FAO inhibition (Figure 5e; Figure S6h, Supporting Information), suggesting that FAO regulates cell death by providing acetyl-CoA.
Next, to examine the relevance of N-alpha-acetylation of CASP2 in the FAO-mediated cell death, we performed a series of experiments.Arrest defective 1 (ARD1) functions as a catalytic subunit of N-acetyltransferase protein complexes and it was reported that knockdown of ARD1 disrupts N-alpha-acetylation of CASP2 induced by DNA damage. [18]We found that knockdown of ARD1 reduced cell death upon DNA damage, consistent with previous work (Figure 5f; Figure S6i, Supporting Information).Importantly, FAO inhibition could no longer affect cell death in ARD1 knockdown cells (Figure 5f).We next explored the direct effects of FAO inhibition on N-alpha-acetylation of CASP2 after DNA damage by performing mass spectrometry analysis.Indeed, we observed that the induction of N-alpha-acetylation of CASP2 after ETS treatment was significantly impaired by FAO inhibition (Figure 5g; Figure S6j, Supporting Information).Finally, we conducted mutagenesis of CASP2 to disrupt its Nalpha-acetylation by replacing the third residue of CASP2 to Pro (A3P) [18] and then expressed WT or mutant CASP2 in cells after knockdown of endogenous CASP2 (Figure S6k, Supporting Information).Consistent with our model, ETO treatment suppressed DNA damage-induced CASP3/7 activities in WT CASP2 expressed cells, whereas it had minimal effects in A3P CASP2 expressed cells (Figure 5h; Figure S6l, Supporting Information).Collectively, these data support the idea that the induction of FAO increased intracellular acetyl-CoA levels after DNA damage, which promotes cell death by enhancing N-alpha-acetylation of CASP2.

Pharmacological Intervention for Improving FAO Rescues Obesity-Induced Chemoresistance
Our results support a model wherein FAO induction is required for DNA damage-induced cell death, which is impaired by obesity.These results led us to hypothesized that restoring FAO CO 2 free incubator at 37 °C for 1 h before the assay.FAO was measured using Seahorse XF Palmitate Oxidation Stress Test Kit.ETO (40 μm) was treated 15 min prior to running assay, and then Palmitate-BSA was added to the wells immediately before running assay.Statistical analysis was performed using one-way ANOVA with Tukey's multiple comparisons test.f,g) Relative PPAR mRNA (f) and protein (g) levels in allograft B16F10 tumors from lean and obese tumors pre-and post-chemotherapy (n = 4).Statistical analysis was performed using two-way ANOVA with Sidak's multiple comparisons test.h) Relative mRNA levels of FAO related genes in B16F10 tumors from lean (up) and obese (down) mice treated with or without WY14643 (40 mg kg −1 five times via intraperitoneal (i.p.) injection) in the presence of ETS (20 mg kg −1 five times via i.p. injection).n = 4 (DMSO, ETS, ETS+WY14643) in WT mice; n = 3 (DMSO, ETS+WY14643) or n = 4 (ETS) in ob/ob mice.Statistical analysis was performed using one-way ANOVA with Dunnett's multiple comparisons test.i) Cell death of B16F10 cells treated with GW6471 (25 μm) or WY14643 (300 μm) and/or ETS (25 μm); n = 3. Statistical analysis was performed using one-way ANOVA with Tukey's multiple comparisons test.j) Tumor volume of B16F10 cells expressing a control shRNA or two independent shRNAs to PPAR in lean mice treated with or without ETS (20 mg kg −1 five times via i.p. injection).Statistical analysis was performed using one-way ANOVA with Tukey's multiple comparisons test.Error bars indicate ±SD (a-e, and i) or ±SEM (f and h).*p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001.; n = 3. Cell death was measured by propidium iodide exclusion assay.g) 293T cells were transfected with FLAG-tagged CASP2 C320G (active cysteine mutant).Transfected cells were treated with or without ETO (200 μm, 24 h) in the presence of ETS (25 μm); n = 4. Lysates were immunoprecipitated with FLAG magnetic beads.The bound proteins were eluted with FLAG peptide and the eluent was performed mass spectrometry.Ratios for levels of N-terminal acetylation were determined by normalizing to total detection peptides.h) Relative CASP3/7 activity of the indicated cells.HeLa cells were infected with mock vector, hCASP2 WT, or hCASP2 A3P and selected by puromycin (2 μg mL −1 ).Infected cells were transfected with siControl or with siRNA against CASP2 (siCASP2).Both cells were treated with or without ETO in the presence of ETS (n = 4).All error bars ±SD.Statistical analysis was performed using one-way ANOVA with Tukey's multiple comparisons test.*p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001.in tumors of obese mice might inhibit their chemoresistance.We used a diet-induced obesity (DIO) mouse model to investigate this issue.After 8 weeks on 60% high-fat diet (HFD), mice weighed 39.1% more than lean, chow diet (CD) control mice (Figure S7a, Supporting Information).To improve FAO induction after chemotherapy, mice subcutaneously injected B16F10 cells were treated with or without WY14643 in the presence of ETS (Figure 6a).First, we tested whether our cellular findings are relevant in this model.In line with our previous results, ETS treatment induced PPAR expression in tumors from control mice, whereas HFD abrogated this induction (Figure 6b).In addition, WY14643 administration strongly induced the expression of PPAR in obese tumors after chemotherapy, whereas it had no further effect on PPAR expression in lean tumors (Figure 6b).Consistent with this, PPAR activation restored the impaired induction of FAO-related genes in obese tumors after chemotherapy (Figure S7b, Supporting Information).We observed comparable results in ob/ob mice (Figure S7c, Supporting Information).Moreover, we demonstrated that ETS treatment increased acetyl-CoA levels in lean tumors but not in obese tumors and PPAR activation could rescue the defect in obese tumors (Figure 6c; Figure S7d, Supporting Information).Thus, consistent with previous results, these data indicate that chemotherapy induces FAO in tumors, leading to increased acetyl-CoA levels, and that obesity impairs this process.
To validate the effect of FAO induction on the sensitivity of tumor cells after chemotherapy, we examined tumors from CDor HFD-fed mice.As expected, chemotherapy-induced cell death was significantly reduced in tumors from HFD-fed mice compared with those from CD-fed mice (Figure 6d).Importantly, the treatment of WY14643 potently synergized with ETS to induce a strong increase of cleaved CASP3 in tumors from HFD-fed mice, resulting in reduced tumor volume in these mice (Figure 6d,e).Consistent with our previous results, WY14643 was able to increase CASP2 activity in obese tumors (Figure 6f; Figure S7e, Supporting Information).However, we found that the combination did not induce the synergistic effects in tumors from control mice (Figure 6d-f; Figure S7e, Supporting Information).Collectively, these results demonstrate that the treatment aimed at improving FAO synergizes with chemotherapy to kill tumor cells in obese mice.

Discussion
In this study, we demonstrate that FAO regulates DNA damageinduced cell death with important implications for obesityinduced chemoresistance.Our study shows that FAO is induced after chemotherapy, which is abrogated by obesity (Figure 1).We find that the induction of FAO is required for apoptotic cell death in response to DNA damage (Figure 2).Molecularly, DNA damage induces FAO via PPAR, which is suppressed by obesity-mediated hypoxia (Figure 3,4).We demonstrate that increased acetyl-CoA production by FAO contributes to DNA damage-induced cell death by promoting N-alpha-acetylation of CASP2, highlighting the potential importance of this pathway in the regulation of chemoresistance of cancer cells (Figure 5).This idea is further validated by the finding that improving FAO reduces obesity-induced chemoresistance (Figure 6).Taken together, our studies provide important insight into the role of FAO as a regulator of DNA damage-induced cell death and uncover a mechanism whereby obesity promotes chemoresistance of tumor cells by inhibiting this critical metabolic response to DNA damage.
Our studies reveal the significant impact of hypoxia on obesitydriven chemoresistance.Hypoxia is a defining feature of solid tumors. [20]Previously, it has been shown that hypoxia leads to resistance to cancer therapies and a higher HIF1 expression is closely associated with poor prognosis in many cancers, including ovarian, colorectal and breast cancers. [21]As inhibiting HIF1 shows efficacy in restoring the decreased response to cancer therapy, it has been proposed that targeting HIF1 has therapeutic potential for efficient cancer treatment. [22]In this study, we demonstrate that obesity aggravates HIF1 stabilization in tumors, inhibiting PPAR-dependent FAO induction after chemotherapy and, thus, suppressing cell death.Importantly, we suggest that the increased FAO is not simply compensative response for maintaining energy homeostasis.Rather, FAO actively functions as an important metabolic regulator of cell death by activating the CASP2-dependent apoptosis pathway.Thus, taken together, these data provide a molecular mechanism whereby hypoxia leads to resistance of cancers to DNA damaging therapies, such as chemotherapy and radiation.
Chemotherapy is designed to kill cancer cells. [23]However, when cancer cells experience non-lethal doses, they can remain to survive or proliferate, or drug-resistant subpopulations can be emerged by clonal expansion. [24]Thus, appropriate chemotherapy doses are crucial for treatment efficacy.Because chemotherapy entails systemic treatment with cytotoxic agents, optimal doses for normal-weight patients are often not enough for overweight or obese patients. [25]Although doses of drugs are determined by using the body surface area, many obese patients received suboptimal chemotherapy dosing, such as due to concerns regarding toxicity, which leads to limited treatment efficacy in obese patients. [26]26b,27] We propose that enhanced FAO significantly attenuates obesity-induced drug resistance.As a proof of concept, we show that PPAR activation synergizes with chemotherapy to induce robust cell death in obese tumors without increasing chemotherapy doses.Thus, our findings may help to develop chemotherapeutic approaches for cancer patients with overweight or obesity.
Although our current work highlights the potential role of FAO induction by PPAR in ameliorating obesity-driven chemoresistance, it does not preclude the involvement of other PPARdependent pathways.As PPAR plays crucial roles in cellular lipid homeostasis, [28] it may regulate DNA damage-induced cell death via other branches of lipid metabolism or other targets.In this regard, fatty acid degradation is closely associated with lipolysis process and lipophagy formation.As these lipid-related pathways lead to the regulation of cell death through lipotoxicity or lipid peroxidation-induced ferroptosis, [29] it will be important for future work to examine how these pathways coordinately regulates cell survival and death upon genotoxic stress.Additionally, it has been reported that DNA damage induces mitochondrial biogenesis through the activation of AMP-activated protein kinase (AMPK). [30]Thus, an increased mitochondrial mass may contribute to the enhanced FAO after DNA damage.On the other hand, we demonstrate that FAO induction in cancer cells contributes to cell death after chemotherapy.However, it is important to note that in vivo cancer tissues consist of several types of cells, including cancer cells, cancer stem cells, inflammatory cells, adipocytes, and various related tissue cells. [31]This heterogeneity of the cancer microenvironment could affect the FAOmediated cell death.In addition, metabolic alteration is a key feature of many cancer cells, which may influence this pathway.For example, metastatic cancer cells exhibit an aberrant dependency on FAO. [32]Previously, it has been shown that FAO inhibition re-sensitize radiation-resistant nasopharyngeal carcinoma cells to radiation-therapy. [33]As therapy-resistant cancer cells acquire resistance, they often undergo metabolic shifts or remodeling, potentially affecting the FAO-dependent DNA damage-induced cell death.Last, there may be cancer-type-specific differences, as different cancer cells have different metabolic dependencies.
This study identifies an essential role of FAO in the DNA damage response by regulating cell death.As the DNA damage response is essential for maintaining cellular and organismal homeostasis, dysregulations of this pathway cause stress-induced premature senescence and are frequently reported in accelerated aging syndromes. [34]We found that cells with the defective FAO induction upon DNA damage resist to cell death.As senescence is an alternative decision for cells when they bypass death, [35] surviving cells can enter a senescence-like state.Intriguingly, recent studies have reported that defects or dysregulations of FAO are observed in senescent cells or in aged tissues. [36]Thus, it is plausible that defects in this metabolic response involving FAO regulation could be one of the causes of the undesirable accumulation of senescent cells during aging.
In sum, our work illustrates that FAO is intimately involved in cellular response to genotoxic stress, in part by regulating DNA damage-induced cell death.Moreover, we found that improving this metabolic response could alleviate obesity-driven chemoresistance.Given the essentiality of metabolism in cell fate decision and cancer progression, these findings may hold profound implications for developing novel therapeutic strategies for cancer as well as for elucidating the cellular metabolic response to stress.
Mice: All animal experiments were performed according to the National Institutes of Health's Guide for the Care and Use of Laboratory Animals, with approval of the Animal Experiment Ethics Committee of the Catholic University of Korea, College of Medicine.4-week-old C57BL/6J and ob/ob mice were purchased from Central Lab Animal Inc (Korea).4week-old C57BL/6 mice were purchased from Orient Bio (Korea).Mice were maintained in a specific pathogen-free facility with 12 h light and 12 h dark cycles.
Mouse Tumor Models: 6-week old C57BL/6J or ob/ob mice were injected subcutaneously in the abdominal flank with 1 × 10 5 B16F10 cells.Tumor volumes were measured using the following ellipsoid formula: volume (mm 3 ) = 0.5 × length × (width 2 ), where length indicates the long diameter and width is the short diameter.When the tumor volumes reached ≈100 mm 3 , ETS (20 mg kg −1 ) and WY14643 (40 mg kg −1 ) were intraperitoneally administered five times everyday.For a diet-induced obesity mice model, 5-week old C57BL/6 mice were fed normal CD or 60% HFD (Research Diets, D12492) for 8 weeks.After 8 weeks of CD or HFD feeding, mice were injected subcutaneously in the abdominal flank with 1 × 10 5 B16F10 cells.When the tumor volumes reached ≈100 mm 3 , ETS (20 mg kg −1 ) and WY14643 (40 mg kg −1 ) were intraperitoneally admin-istered five times everyday.A day after a course of treatment, mice were sacrificed for collection of tumor tissue.
RNA Extraction and Quantitative RT-PCR: RNA was isolated from cell lines or tumor tissues using Trizol (Invitrogen, 15596026) according to the manufacturer's instructions.0.5 μg of total RNA was reverse-transcribed using the PrimeScript RT Master Mix (Takara, RR036A).Quantitative RT-PCR was performed using SYBR Green I Master (Roche, 04887352001) on a Lightcycler 480 (Roche).The level of gene expression was normalized to -actin.The primer sequences used for qPCR are listed in Supplementary Table (Supporting Information).
Western Blotting: Cells were lysed with EZ-RIPA lysis buffer (ATTO, WSE-7420) supplemented with protease inhibitor cocktail (Roche, 11873580001).Cell lysates were separated by SDS-PAGE and transferred onto nitrocellulose membranes (GE Healthcare) or polyvinylidene fluoride membranes (GE Healthcare).The membranes were blocked for 1 h in 5% nonfat milk or 3% BSA and then incubated with primary antibodies overnight at 4 °C.After washing five times in 1X TBST, the membranes were incubated with species-specific secondary antibodies and detected using LAS 4000.
Flow Cytometric Measurement of Cell Death: Cells at less than 70% confluence were treated with drugs.After treatment, cells were harvested by trypsinization, pelleted by centrifugation, and resuspended in PBS.The measurement of cell death was performed by flow cytometry using propidium iodide (PI) staining and/or annexin V staining (BD Biosciences, 556547).
Caspase 3/7 GLO: Cells were plated into 96-well plates at 2000 cells per well in 100 μL of growth media.Casp2 WT and CASP2 mutant (A3P) stable cell lines were transfected with siCASP2 for 24 h and plated into 96well plates at 2000 cells per well in 100 μL of growth media.The following day, cells were treated with ETS, ETO, or both.100 μL of Caspase-GLO 3/7 reagent (Promega, G8091) were added to each well of white-walled 96-well plate and incubated at room temperature for 30 min.After incubation, the luminescent signal was detected using plate-reading luminometer.

Figure 1 .
Figure 1.Impairment of FAO induction is related to chemoresistance in obesity.a) Immunohistochemistry of cleaved CASP3 expression pre-and postchemotherapy in B16F10 tumors from lean and obese mice.Scale bar represents 50 μm.b) Number of genes related to top 10 KEGG pathways.c) Enrichment of KEGG metabolic signature scores that are altered by chemotherapy in RNA-seq from lean (yellow-green bars) and obese tumors (blue bars).d) Volcano plots of differentially expressed fatty acid oxidation genes in tumors from lean (upper panel) and obese mice (lower panel) post-and pre-chemotherapy (>1.5-fold, p < 0.05).e) Schematic representation of mitochondrial fatty acid oxidation.f) Relative expression of FAO related genes in B16F10 tumor cells from lean (upper panel) and obese mice (lower panel); n = 4. Statistical analysis was based on two-tailed Student's t-test.All error bars ±SEM.*p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001.

Figure 3 .
Figure 3. PPAR is an upstream target for DNA damage-induced FAO induction.a-c) Relative Ppar mRNA (a), protein (b), and transcriptional activity (c) levels in B16F10 cells treated with ETS (25 μm, 24 h); n = 3. Statistical analysis was performed using two-tailed Student's t-test.d) Relative expression of FAO related genes in B16F10 cells treated with GW6471 (25 μm, 24 h) or WY14643 (300 μm, 24 h) in the presence of ETS (25 μm, 24 h); n = 4. Statistical analysis was performed using one-way ANOVA with Tukey's multiple comparisons test.e) FAO-dependent ΔOCR of B16F10 cells treated with GW6471 (25 μm) or WY14643 (300 μm) and/or ETS (25 μm); n = 5.The cell medium was replaced with assay media and the cells were incubated in a

Figure 4 .
Figure 4. Obesity-mediated hypoxia inhibits the FAO induction after chemotherapy.a) Immunohistochemistry of HIF1 expression in lean and obese B16F10 tumors.Scale bars represent 50 μm.Statistical analysis was performed using two-tailed Student's t-test.Error bars represent ±SEM b) Relative Ppar mRNA levels in B16F10 cells treated with ETS (25 μm) under normoxic or hypoxic conditions for 24 h (n = 3).Statistical analysis was performed using one-way ANOVA with Tukey's multiple comparisons test.c,d) Relative mRNA (c) and protein (d) levels of FAO related genes in B16F10 cells treated with ETS (25 μm) under normoxic or hypoxic conditions for 24 h (n = 3).Statistical analysis was performed using two-way ANOVA with Sidak's multiple comparisons test.e) HIF1 and PPAR expression of B16F10 cells transfected with nontargeting siRNA (siControl) or with siRNA against HIF1 (siHIF1) under hypoxic condition.-actin serves as a loading control.f) Relative expression of FAO related genes in B16F10 cells transfected with nontargeting siRNA (siControl) or with siRNA against HIF1 (siHIF1) under hypoxic condition (n = 3).Statistical analysis was performed using one-way ANOVA with Tukey's multiple comparisons test.Error bars ±SD (b-f).*p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001.

Figure
Figure FAO adjusts DNA damage-induced cell death by N-alpha-acetylation of CASP2.a) Cell death of B16F10 cells transfected with nontargeting siRNA (siControl) or with siRNA against P53 (siP53).Cells were treated with ETS (25 μm, 40 h), ETO (50 or 200 μm, 40 h) or both (n = 3).b) Cleaved CASP2 expression of 293T cells treated with or without ETO in the presence of ETS.-actin serves as a loading control.c) Cell death of B16F10 cells transfected with nontargeting siRNA (siControl) or with siRNA against CASP2 (siCASP2).Cells were treated with or without ETO (50 μm, 40 h) in the presence of ETS (25 μm, 40 h); n = 3. Cell death was measured by propidium iodide exclusion assay.d) Relative acetyl-CoA levels were measured in immortalized MEFs treated with ETO (200 μm, overnight) and/or citrate (10 mm, overnight) in the presence of ETS (25 μm, overnight); n = 4. e) Cell death of immortalized MEFs treated with the indicated drugs (n = 3).Cell death was measured by propidium iodide exclusion assay.f) Cell death of immortalized MEFs transfected with nontargeting siRNA (siControl) or with siRNA against ARD1 (siARD1).Cells were treated with or without ETO (200 μm, overnight) in the presence of ETS (25 μm, overnight); n = 3. Cell death was measured by propidium iodide exclusion assay.g) 293T cells were transfected with FLAG-tagged CASP2 C320G (active cysteine mutant).Transfected cells were treated with or without ETO (200 μm, 24 h) in the presence of ETS (25 μm); n = 4. Lysates were immunoprecipitated with FLAG magnetic beads.The bound proteins were eluted with FLAG peptide and the eluent was performed mass spectrometry.Ratios for levels of N-terminal acetylation were determined by normalizing to total detection peptides.h) Relative CASP3/7 activity of the indicated cells.HeLa cells were infected with mock vector, hCASP2 WT, or hCASP2 A3P and selected by puromycin (2 μg mL −1 ).Infected cells were transfected with siControl or with siRNA against CASP2 (siCASP2).Both cells were treated with or without ETO in the presence of ETS (n = 4).All error bars ±SD.Statistical analysis was performed using one-way ANOVA with Tukey's multiple comparisons test.*p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001.

Figure 6 .
Figure 6.PPAR agonist improves the obesity-induced chemoresistance.a) Schematic depicting in vivo experimental setup.b,c) b) Relative Ppar mRNA (n = 7 for CD and n = 10 for HFD) and c) acetyl-CoA (n = 5) levels in allograft B16F10 tumors from CD-and HFD-fed mice treated with or without WY14643 (40 mg kg −1 five times via i.p. injection) in the presence of ETS (20 mg kg −1 five times via i.p. injection).Statistical analysis was performed using two-way ANOVA with Tukey's multiple comparisons test.d) Immunohistochemistry of cleaved CASP3 expression in lean and obese tumors treated with or without WY14643 in the presence of ETS.Scale bar represents 50 μm.Statistical analysis was performed using one-way ANOVA with Tukey's multiple comparisons test.e) Tumor volume of B16F10 tumors in CD-and HFD-fed mice treated with or without WY14643 in the presence of ETS.f) Immunoblot for cleaved CASP2 in lean and obese tumors treated with or without WY14643 in the presence of ETS.All error bars ±SEM.**p < 0.01, ***p < 0.001, and ****p < 0.0001.