Isoprenylcysteine Carboxylmethyltransferase Regulates Mitochondrial Respiration and Cancer Cell Metabolism

Home , ICMT, Prenylation

ORIGINAL ARTICLE Isoprenylcysteine carboxylmethyltransferase regulates mitochondrial respiration and cancer cell metabolism

JT Teh1,5, WL Zhu1,2,5, OR Ilkayeva3,YLi4, J Gooding3, PJ Casey1, SA Summers4, CB Newgard3 and M Wang1,2

Isoprenylcysteine carboxylmethyltransferase (Icmt) catalyzes the last of the three-step posttranslational protein prenylation process for the so-called CaaX proteins, which includes many signaling proteins, such as most small GTPases. Despite extensive studies on Icmt and its regulation of cell functions, the mechanisms of much of the impact of Icmt on cellular functions remain unclear. Our recent studies demonstrated that suppression of Icmt results in induction of autophagy, inhibition of cell growth and inhibition of proliferation in various cancer cell types, prompting this investigation of potential metabolic regulation by Icmt. We report here the ﬁndings that Icmt inhibition reduces the function of mitochondrial oxidative phosphorylation in multiple cancer cell lines. In-depth oximetry analysis demonstrated that functions of mitochondrial complex I, II and III are subject to Icmt regulation. Consistently, Icmt inhibition decreased cellular ATP and depleted critical tricarboxylic acid cycle metabolites, leading to suppression of cell anabolism and growth, and marked autophagy. Several different approaches demonstrated that the impact of Icmt inhibition on cell proliferation and viability was largely mediated by its effect on mitochondrial respiration. This previously unappreciated function of Icmt, which can be therapeutically exploited, likely has a signiﬁcant role in the impact of Icmt on tumorigenic processes.

Oncogene (2015) 34, 3296–3304; doi:10.1038/onc.2014.260; published online 25 August 2014

INTRODUCTION These phenotypes suggested to us that the cells in which Icmt is Prenylation is a three-step posttranslational lipid modification inhibited might be metabolically compromised, as impairment of process in the maturation of a number of proteins involved in cell ATP production induces a starvation response leading to similar regulation. The majority of prenylated proteins belong to a group presentations. We therefore investigated the impact of Icmt termed ‘CaaX proteins’ that are defined by a specific C-terminal inhibition on cell respiration, energy status and metabolism. Our amino-acid sequence ‘cysteine-aliphatic-aliphatic-any’, serving as findings demonstrate that Icmt has a previously unappreciated fi the consensus sequence for their modification. Following role in cellular metabolism, which account signi cantly for its isoprenoid addition on the cysteine and proteolytic removal of impact on growth and proliferation and its role in tumorigenesis. the –aaX sequence, the final step is methylation of the C-terminal prenylcysteine by isoprenylcysteine carboxylmethyltransferase (Icmt).1,2 The prenylation process is required for proper function RESULTS of the modified protein, either as a mediator of specific subcellular Icmt inhibition leads to AMPK activation, a result of energy localization, a determinant for specific protein–protein interac- depletion 3–5 tions, or protein stability. Following the discovery that K-Ras We have shown previously that treatment of multiple cancer cell is significantly mislocalized in cells that lack the Icmt types, including PC3 prostate and MDA-MB-231 breast cancer 4 methyltransferase, and that the abilities of K-Ras to transform cells, with a small-molecule Icmt inhibitor termed cysmethynil led fibroblasts and promote myeloproliferative disease and lung to inhibition of the mammalian target of rapamycin (mTOR) and cancer are dependent on Icmt,6,7 this enzyme has gained elevated autophagy.8,9 This Icmt inhibitor-induced autophagy attention as a potential cancer target. However, it is increasingly phenotype was extensively investigated in our prior studies, which recognized that the impact of Icmt-catalyzed CAAX protein demonstrated consistent increased LC3 I to LC3 II conversion and methylation is not limited to processes mediated by Ras; this increased autophagy flux measured by multiple modalities, such carboxylmethylation also affects basic cell functions mediated by as baflomycin cotreatment and colocalization of RFP-LC3- and CaaX proteins other than Ras.2 Unravelling the roles of Icmt in GFP-LC3-positive vesicles. However, the mechanism through regulating CaaX protein-driven processes is fundamentally which Icmt inhibition impacts this important biologic response important. was unresolved. To investigate the impact of cysmethynil Pharmacologic or genetic suppression of Icmt results in slow treatment upstream of mTOR, we assessed the activation status cell growth, cell cycle arrest and excessive autophagy, of possible regulators including AMP-dependent kinase (AMPK). which account for the anticancer efficacies of Icmt inhibition.8,9 AMPK activity responds to ATP levels and hence provides a direct

1Program in Cancer and Stem Cell Biology, Duke-NUS Graduate Medical School, Singapore, Singapore; 2Department of Biochemistry, National University of Singapore, Singapore, Singapore; 3Sarah W Stedman Nutrition and Metabolism Center, and Duke Institute of Molecular Physiology, Duke University Medical Center, Durham, NC, USA and 4Program in Cardiovascular and Metabolic Disorders, Duke-NUS Graduate Medical School, Singapore, Singapore. Correspondence: Dr M Wang, Program in Cancer and Stem Cell Biology, Duke-NUS Graduate Medical School, 8 College Road, Singapore 169857, Singapore. E-mail: [email protected] 5These authors contributed equally to this work. Received 4 February 2014; revised 9 June 2014; accepted 3 July 2014; published online 25 August 2014 Icmt regulates mitochondrial respiration JT Teh et al 3297 gauge of cellular energy status. Once activated, AMPK suppresses cysmethynil-treated PC3 cells (Figure 1d). Although ATP is anabolic activities, activates fuel catabolism and promotes considered the major energy currency in cells, it is important to autophagy to increase energy stores.10,11 measure the other nucleoside triphosphates in the evaluation of Indeed, the treatment of PC3 cells with cysmethynil elevated cell energy status, as there exists a dynamic balance between levels of phosphorylated AMPK in a dose-dependent manner different NTPs, energy currency molecules.12,13 The reduction of (Figure 1a). We also observed increased inhibitory phosphoryla- NTPs in cysmethynil-treated cells suggests that AMPK activation is tion of acetyl-coA carboxylase (ACC) at serine 79, an AMPK likely the result of energy deficiency. It is worth noting that the phosphorylation site, consistent with increased AMPK activity. energy deplete state and associated signaling changes induced by Aligned with our previous reports of autophagy induction, there Icmt inhibition is not limited to PC3 cells; MDA-MB-231 cells was also a dose-dependent accumulation of the autophagy responded in a similar manner (Supplementary Figure 1), marker LC3 II, which paralleled the response of AMPK activation suggesting a general regulatory mechanism by Icmt in cell energy (Figure 1a). Similar phenotype of elevated pAMPK level was metabolism. observed upon small interfering RNA (siRNA) suppression of Icmt expression (Figure 1b). Further, we subjected wild-type mouse Cysmethynil treatment reduces mitochondrial respiratory capacity embryonic fibroblast (MEF) cells and Icmt-null MEF cells to To investigate the cause of cysmethynil-induced energy depletion, cysmethynil treatment. While Icmt-null MEFs exhibit higher basal we studied mitochondrial function in the cells. Cysmethynil- pAMPK and LC3 II levels, robust increases in pAMPK and LC3 II treated cells exhibit markedly reduced basal/resting-respiration levels are only observed in the wild-type MEFs upon cysmethynil rates (Figure 2a). In addition, cysmethynil-treated cells exhibited a treatment (Figure 1c). These genetic suppression studies provided shallower drop in oxygen consumption rate (OCR) upon the compelling evidence that effect of cysmethynil on AMPK addition of oligomycin, an ATPase inhibitor, in comparison with activation and autophagy is Icmt dependent. the untreated cells (Figure 2a), suggesting attenuated ATP The parallel elevation of autophagy and activation of AMPK are production before oligomycin addition. This result is consistent indications that Icmt-induced autophagy is mediated by AMPK with the NTP quantification study and the phenotype of AMPK activation, which can be either a direct response to cellular energy activation shown above. Trifluorocarbonylcyanide phenylhydra- status or a result of modulation by upstream molecules. In the zone (FCCP), uncoupling the electron transport system from investigation for possible etiology of Icmt-inhibition-mediated oxidative phosphorylation, is often used to assess the maximal AMPK activation, we analyzed the levels of nucleotide tripho- respiratory capacity of cells. Cysmethynil-treated cells displayed sphates (NTPs) in cysmethynil-treated and control cells. Significant lower FCCP-induced respiration in comparison with untreated decreases in levels of ATP, GTP, CTP and UTP were observed in cells (Figure 2a), indicative of reduction in the maximal respiratory capacity and potential rate for ATP production. Last, the differences in respiration between cysmethynil-treated and control cells are not likely from non-mitochondrial oxygen

PC3 1.2 ﬁ Cysm consumption, as no signi cant differences of OCRs relative to 0.9 the basal respirations were observed between these two cell p-AMPK (T127) 0.6 populations after the addition of rotenone and antimycin A p-ACC (Ser79) p-AMPK (T127) Expression (Figure 2a). The table below panel a summarizes the OCR values, 0.3 LC3 I/II LC3 I /II illustrating the reduction in mitochondrial respiratory capacity and

ICMT 0.0 mitochondrial ATP production, under the treatment of GAPDH GAPDH cysmethynil. To investigate whether the defect(s) in energy production in response to Icmt inhibition is at a stage upstream of mitochondria, we measured media acidiﬁcation as a surrogate for glycolytic Cysm (µM) 0 20 22.5 2.5 lactate production for PC3 cells. Upon glucose addition, Icmt +/+ -/- +/+ -/- +/+ -/- 2.0 cysmethynil-treated cells demonstrated a markedly higher rate cells

6 fi p-AMPK (T127) 1.5 of acidi cation of media as compared with that of untreated cells 1.0 (Figure 2b), indicating a higher rate of glycolysis. This observation LC3 I/II 0.5 suggests that the cells are poised to use glucose via anaerobic nmol/ 10 GAPDH 0.0 glycolysis, and that neither glucose transport nor the glycolytic ATP GTP CTP UTP pathway is compromised by Icmt inhibition. Subsequently, ATP Figure 1. ICMT inhibition leads to AMPK activation and reduction of synthesis inhibitor oligomycin was added to assess maximal nucleoside triphosphates. (a) AMPK activation and autophagy glycolytic capacity of the cells. Upon addition of oligomycin, induction in cysmethynil-treated PC3 cells. Cells were treated with control cells showed a significant increase in extracellular increasing doses of cysmethynil—0 (dimethyl sulfoxide (DMSO)), fi 8 acidi cation rate (ECAR), whereas cysmethynil-treated cells 17.5, 20 and 22.5 μM—for 24 h as described before preparation of b exhibited no additional increase (Figure 2b). There was no cell lysates and immunoblot analysis. ( ) Suppression of Icmt fi expression leads to AMPK activation. Left, pAMPK and LC3 levels in signi cant difference in the maximal glycolytic capacity between PC3 cell lysates 96 h after transfection of siRNAs (Invitrogen) these two populations of cells, although the maximal glycolysis targeting Icmt or luciferase (as a control); right, analysis of Icmt was achieved with the addition of glucose in the cysmethynil- knockdown efficiency by RT–PCR. (c) Immunoblot of pAMPK and treated cells, whereas in the control cells this required the addition LC3 in lysates from Icmt wild-type MEF (+/+) and Icmt-null MEF of oligomycin (Figure 2b). These observations suggest that Icmt ( − / − ) cells (obtained from M Bergo) following 24 h of treatment inhibition did not compromise glycolysis; hence, the reduction of with either DMSO or the indicated concentration of cysmethynil. OCR and energy depletion is the results of cysmethynil-induced d ( ) Quantitative analysis of nucleoside triphosphates in control and dysfunction of mitochondria oxidative phosphorylation. The table cysmethynil-treated PC3 cells. The cells were treated with DMSO μ μ in Figure 2b presents the calculated values of glycolytic flux and (black bar), 20.0 M cysmethynil (gray bar) or 22.5 M cysmethynil fi (white bar) for 24 h before being harvested and subjected to capacity from the ECAR studies. Similar ndings of OCR and ECAR nucleotide analysis as described.34 Error bars denote standard changes had been observed in cysmethynil-treated MDA-MB-231 deviation of the data from three technical repeats. A biologic repeat cells, illustrating that this regulation by Icmt is not cell line specific of the study exhibited similar results. (Supplementary Figure 2).

15 300

10 200

5 ECAR (%) 100

0 0 OCR (pmol O2/ min/ µg)min/ O2/ OCR (pmol 0255075100125 0 22446688110 Time (minutes) Time (minutes)

OCR DMSO Cysm ECAR DMSO Cysm Max. Resp. 6.77±0.99 3.52±0.23 ± ± Capacity Glycoly. Flux 0.68 0.03 1.09 0.25 ATP Prod. 3.51±0.51 2.44±0.26 Glycoly. Capacity 1.25±0.07 1.09±0.21

Ant Olig FCCP & Rot 40 1.5 si-Luc si-ATG5 30 1.0 Cysm – + – + 20 ATG5 0.5 10

Mito. [DNA] Mito. LC3 I/II 0

0.0 ug) min/ OCR (pmol/ GAPDH cysm 0 306090120 Time (minutes)

Figure 2. Cysmethynil treatment targets mitochondrial respiration. OCR analysis (a) and glycolysis flux analysis measured by ECARs (b). PC3 cells were treated with either dimethyl sulfoxide (DMSO) (black tracing) or 20 μM cysmethynil (gray tracing) for 24 h before the studies. One hour before performing the OCR or ECAR assays, the culture media were replaced with prewarmed XF assay medium (Seahorse Bioscience) for each study, respectively, and incubated at 37 1C according to Seahorse Bioscience protocols. OCRs were measured under basal condition first, followed by that after sequential addition of oligomycin, FCCP, and antimycin A and rotenone. ECAR tracings were obtained at baseline, and then after glucose injection, addition of oligomycin and injection of 2-deoxy-D-glucose (2-DG), sequentially. The tables in the lower part of (a) and (b) summarize the calculated mitochondrial respiratory capacity and ATP production rate, and glycolytic rates upon glucose addition and glycolytic capacities, respectively. All the OCR and ECAR values presented were normalized to protein mass. Error bars represents standard errors of the means. All experiments were repeated more than two times with similar results. (c) Mitochondrial DNA quantification of PC3 cells after the treatment with DMSO (black bar), 20 μM (gray bar) or 22.5 μM (white bar) of cysmethynil for 24 h. Mitochondrial DNA quantity was normalized to that of genomic DNA. Error bars indicate standard deviations of three separate measurements. (d) Immunoblot analysis of Atg5 and LC3 levels in MDA-MB-231 cells subject to Atg5 knockdown followed by DMSO ( − ) or 20.0 μM cysmethynil (+) treatment. (e) Mitochondria respiration analysis of MDA-MB-231 cells treated with DMSO (black tracing) or 20 μM cysmethynil (red tracing) following Atg5 knockdown. The cells were transfected with Atg5 siRNA (Invitrogen), and 48 h later subjected to DMSO or cysmethynil treatment for 24 h, whereupon the respiration rates of the cells were measured at basal conditions and then after the addition of oligomycin, FCCP and rotenone+antimycin A as indicated.

The ﬁndings that Icmt inhibition leads to mitochondrial rescue of mitochondria dysfunction was observed (Figure 2e). dysfunction and an energy-depleted state in the cells, coupled Indeed, the reduction of mitochondrial respiration induced by with the previous reports that Icmt inhibition induces cysmethynil treatment was as pronounced in the Atg5- autophagy,8,9 raise the need to clarify the sequence of events: knockdown cells as in the parental cells (Figure 2a). Consistently, mitochondrial dysfunction and the ensuing energy depletion as similar results were obtained with Atg7-knockdown cells the cause for autophagy response, or excessive autophagy/ (Supplementary Figure 3). These data from different approaches mitophagy as the cause for mitochondria depletion and provided compelling evidence to exclude mitophagy as a cause subsequent dysfunction. Therefore, we investigated the role of for cysmethynil-induced mitochondrial dysfunction and to autophagy in the mitochondrial respiration changes induced by strengthen the notion that autophagy is likely the response to Icmt inhibition using two different approaches. Quantitative PCR energy depletion. study demonstrated that cysmethynil-treated PC3 cells had similar amount of mitochondrial DNA compared with that of control cells Cysmethynil treatment reduces mitochondrial complex I, II and III (Figure 2c), suggesting that reduction of mitochondria quantity activities could not account for the decreased oxygen consumption. Direct We next focus on Icmt regulation of mitochondrial respiratory support for mitochondria dysfunction occurring before or function. The main function of mitochondrial complex I and independent of autophagy came from examination of the impact complex II are extracting electrons from their respective of concurrent cysmethynil treatment and siRNA knockdown substrates. Thus, we examined mitochondrial oxidations of of Atg5 and Atg7, proteins essential for autophagy,14,15 on complex I-associated substrate and complex II substrate in mitochondrial function. While siRNA knockdown of Atg5 com- cysmethynil-treated cells by high-resolution in vivo oximetry, pletely abolished cysmethynil-induced autophagy (Figure 2d), no measuring real-time oxygen consumption. The tracings for oxygen

Oncogene (2015) 3296 – 3304 © 2015 Macmillan Publishers Limited Icmt regulates mitochondrial respiration JT Teh et al 3299 concentration and oxygen consumption of PC3 cells subjected to assess complex II activity after the irreversible inhibition of the addition of various substrates and inhibitors are shown in complex I by rotenone. Cytochrome c was used for the assessment Figure 3a. Basal respiration rates are first measured at resting state, of mitochondrial membrane integrity. These studies demonstrated after which the plasma membrane is selectively permeabilized that treatment of PC3 with cysmethynil lowered basal respiration with digitonin for subsequent studies. Peak function of complex I and complex I and II activities, as shown in Figure 3a tracing. To is measured by the addition of malate and pyruvate (substrates of further dissect mitochondrial electron transport system, we complex I), and ADP. Complex I-associated respiration is measured measured oxygen consumptions triggered by duroquinol as the drop, ΔI, from this peak, induced by the addition of and ascorbate–N,N,N′,N′-tetramethyl-p-phenylenediamine (TMPD), rotenone, an inhibitor of complex I. Subsequently, complex which are specific artificial electron donors for complexes III II-associated respiration (ΔII) was measured by determining the and IV, respectively.16 Similar to the method used for complexes difference between succinate-induced respiration and the residual I and II, complex III and IV activities (ΔIII and ΔIV) were obtained respiration after the addition of malonate or antimycin (complex II by subtracting the residual respiration in the presence of and III inhibitors, respectively). As both complex I and II antimycin (complex III inhibitor) or KCN (complex IV inhibitor), respirations are mediated through complex III electron transport, from peak respiration triggered by duroquinol or ascorbate– either malonate or antimycin after addition of the succinate can TMPD, respectively. It was observed that duroquinol-stimulated

PYR+MAL ROT 120 DIG ADP Cyt c SUC ANT 100 80 250 100 DMSO 60 200 80

OCR (%) 40 150 60 20 100 40 ΔII 0 Δ Basal ΔI ΔII ΔIII ΔIV 50 I 20 cells) 6

0 0 24 h 48 h 250 100 Cysm(µM) - + - + Cysm 200 80 NDUFS3 (C I) [O2 ] (nmol/ mL) 150 60 SDHA (C II) OCR (pmol/s/ 10 100 40 UQCRC2 (C III) Δ COXIV (C IV) 50 Δ II 20 I ATP5A (C V) 0 0 VDAC 0 1020304050 GAPDH Time

Complex I Complex II Complex IV Complex V 0.308 0.48 0.11 0.372 0.303 0.10 0.369 0.46 0.09 0.366 0.298 OD 340 OD 550

OD 600 0.44 OD 450 0.08 0.363 0.293

0.07 0.36 0.288 0.42 120 360 600 960 1210 1460 420 930 1440 120 540 960 Time (sec) Time (sec) Time (sec) Time (sec) Figure 3. Cysmethynil treatment leads to reduction of mitochondrial complex I and complex II activities, in Icmt-dependent manner. (a) Oximetry profiles of PC3 cells subjected to either dimethyl sulfoxide (DMSO) control (upper panel) or 22.5 μM cysmethynil (lower panel) treatment for 24 h. Respiratory oxygen consumption was assessed in real-time using an Oroboros Oxygraph-2k respirometer. Oxygen levels (left vertical axis; blue tracing) and real-time oxygen consumption rates (right vertical axis; red tracing) were obtained for intact (basal) cells, followed by that after treatment with the permeabilization agent digitonin, and then measured after the treatment with the indicated series of electron transport chain (ETC) substrates and inhibitors, in the order of pyruvate+malate, ADP, cytochrome c, rotenone, succinate and antimycin A. Mitochondrial complex I-associated substrate oxidation (ΔI) is the change between pyruvate- and malate-stimulated respiration and the residual oxygen consumption after injection of rotenone; substrate oxidation of mitochondrial complex II (ΔII) was measured by the difference in respiration between succinate-induced respiration and residual oxygen consumption after antimycin inhibition. (b) Basal respiration, complex I and complex II substrate oxidations (ΔI and ΔII), duroquinol-triggered respiration (ΔIII) and ascorbate–TMPD-stimulated complex IV respiration (ΔIV), following 24 h treatment with DMSO vehicle control (black bar), 20.0 μM (gray bar) and 22.5 μM cysmethynil (white bar). Data represent the mean and standard deviation of biologic triplicate. (c) Immunoblot quantification of mitochondrial complex subunits NADH dehydrogenase (ubiquinone) iron-sulfur protein 3 (NDUFS3; complex I), succinate dehydrogenase complex subunit A (SDHA; complex II), cytochrome b–c1 complex subunit 2 (UQCRC2; complex III), cytochrome c oxidase subunit IV (COXIV; complex IV) and mitochondrial F1 complex alpha polypeptide (ATP5A; complex V) in cysmethynil- (20 μM) treated PC3 cells, at 24 and 48 h, in comparison with vehicle control. Voltage-dependent anion channel (VDAC) and glyceraldehyde 3-phosphate dehydrogenase (GAPDH) serve as loading control. (d) In vitro enzymatic activity profiles of mitochondrial complex I, II, IV and V, after 24 h treatment with 20.0 μM (gray square) and 22.5 μM (gray triangle) cysmethynil, in comparison with DMSO control (black square). Graphs are from single experiment that has been repeated two times with similar results.

© 2015 Macmillan Publishers Limited Oncogene (2015) 3296 – 3304 Icmt regulates mitochondrial respiration JT Teh et al 3300 respiration (complex III) decreases with cysmethynil treatment, as readout, oxidations of site I and site II substrates depend not whereas ascorbate–TMPD-stimulated respiration (complex IV) only on complexes I and II but also on the distal complexes III and remained unchanged. The high-resolution oximetry studies have IV. Even though arguments can be made for substrate- concluded that Icmt inhibition suppresses complex I, II and III independent reduction of complex I, II and III activities based on function, while spares that of complex IV, in the electron transport evidences from western blot and the duroquinol and ascorbate- chain (Figure 3b). Similar results were obtained when the –TMPD respiration studies, direct measurement of enzymatic experiments were performed in MDA-MB-231 cells (Supplementary activities of each complex can provide more definitive conclusion. Figure 4A). To this end, complex I, II, IV and V enzymatic assays were In support of oximetry studies, immunoblot analysis of a panel performed after extraction and immunoprecipitation of functional of proteins, containing a subunit from each mitochondrial complexes. These assays demonstrate that cysmethynil treatment complex, has shown consistent changes in cysmethynil-treated reduced complex I and II activities but not that of complex IV and PC3 cells. As compared with the control, decreases in complex I, II V activities (Figure 3d), consistent with the results from oximetry and III subunits were observed after cysmethynil treatment; these and immunoblot studies. proteins are NADH dehydrogenase (ubiquinone) iron-sulfur In summary, our evidence suggests that Icmt regulates protein 3 (NDUFS3), succinate dehydrogenase complex subunit mitochondrial respiration, specifically functions of complex I, II A (SDHA) and cytochrome b–c1 complex subunit 2 (UQCRC2) from and III; inhibition of Icmt results in reduction in the function of complex I, II and III, respectively. As expected, there is no these complexes, leading to reduction of mitochondrial respira- significant quantitative changes of cytochrome c oxidase subunit tion and development of an energy-depleted state. IV (COXIV) and ATP synthase, H+ transporting, mitochondrial F1 complex alpha polypeptide (ATP5A), complex IV and V subunit, Cysmethynil-induced suppression of OXPHOS complex activities is respectively (Figure 3c). Of note, same immunoblot observations ICMT specific were made in MDA-MB-231 cells (Supplementary Figure 4B). In To investigate the target specificity of cysmethynil in suppressing addition to support the conclusion of oximetry study, this analysis mitochondrial respiratory function, oximetry studies were also also indicates that cysmethynil impacts on the quantities of the carried out in wild-type MEF and Icmt-null MEF cells. Icmt-null affected complexes, the likely cause for decreased function of MEFs showed no significant changes in respiration upon the same. cysmethynil treatment. In contrast, wild-type MEFs responded to It is important to note that oximetry study with intact the Icmt inhibition with reduced resting- respiration, substrate- mitochondria, as described above, comes with a few limitations. induced complex I and II respiration and duroquinol-stimulated First, the assay does not eliminate substrate availability as a factor complex III respiration (Figures 4a and b), similar to that observed of mitochondrial respiration. Even though saturated substrates are in PC3 and MDA-MB-231 cells. The calculated changes in OCR are provided during the assay, differences in other tricarboxylic acid summarized in Figure 4c. The resistance of Icmt-null MEF to cycle intermediates, enzymes and transporters or carriers can be cysmethynil-induced mitochondrial dysfunction provides strong contributing factors for the decreased oxidation observed. evidence that this consequence of cysmethynil treatment is Icmt Second, as oxygen consumption (complex IV function) was used dependent.

Wild Type MEF Icmt-null MEF DIG PYR + MAL + ADP 190 190 PYR + MAL + ADP DIG DQH 180 ROT ROT DQH ANT 185 SUC SUC ANT 170 MLN MLN 180 160 175 150 [O2] (nmol/ml) [O2] (nmol/ml) 140 170

130 165 0 1020304050 0 1020304050 Time (Min) Time (Min)

Change in OCR (%) Wild Type MEF Icmt null MEF p-value Basal -26.3 ± 18.4 + 4.0 ± 9.8 0.0177 ΔI -60.3 ± 13.3 -10.2 ± 12.5 0.0002 ΔII -64.2 ± 16.0 -2.8 ± 19.37 0.0003 ΔIII -56.9 ± 15.9 + 11.5 ± 15.9 0.0006

Figure 4. Cysmethynil induced inhibition of respiratory complexes is ICMT dependent. (a) and (b) illustrate oxygen consumption rate by Oroboros oximetry analysis of wild type MEF cells and Icmt-null MEF cells, respectively, following 24 h of either dimethyl sulfoxide (DMSO) (black tracing) or 20.0 μM cysmethynil (gray tracing) treatments. (c) Calculated changes of basal respiration, complex I and complex II substrate oxidations (ΔI and ΔII), and duroquinol-triggered respiration (ΔIII) of wild-type and Icmt-null MEF cells following cysmethynil treatment. Data shown represent mean and standard deviation of data pooled from ﬁve independent experiments.

Oncogene (2015) 3296 – 3304 © 2015 Macmillan Publishers Limited Icmt regulates mitochondrial respiration JT Teh et al 3301 Cysmethynil treatment leads to reduction of critical metabolites metabolite depletion, induced by Icmt inhibition, and the and cellular anabolic activity expression of these anabolic enzymes. In addition to regulating It is recognized that compromised oxidative phosphorylation and the expression levels of these proteins, AMPK also suppresses the the ensuing energy deficiency exert significant impact on cell activity of ACC and 3-hydroxy-3-methylglutaryl-CoA reductase 19,20 metabolism. To investigate the metabolic alterations, tricarboxylic through direct phosphorylation. Indeed, this is consistent with acid cycle intermediates were quantified in cysmethynil-treated increased pACC levels in cysmethynil-treated cells (see Figure 1a). and control MDA-MB-231 cells; levels of most of these were Investigating the impact of Icmt inhibition on key modulators of significantly reduced in cysmethynil-treated cells (Figure 5a). protein synthesis, an important component of cell anabolism, we Similar organic acid profiles are observed in control and observed mTOR pathway suppression with the characteristic cysmethynil-treated PC3 cells (Supplementary Figure 5A). Inter- pattern of ‘downward-shift’ of phospho-4EBP1 on immunoblot estingly, amino-acid profiling demonstrated no reduction in most (Figure 5f). 4EBP1 is an archetypical downstream effector of mTOR of the amino acids, but a significant reduction in aspartate and in regulating protein translation; as in the synthesis of lipids, AMPK asparagine (Asx) and a less robust reduction of glutamate and also has an central role in regulating mTOR signaling in response 21,22 glutamine levels (Glx) (Figure 5b). The reduction of Asx and Glx to cellular energy status. The changes of phospho-4EBP1 levels could be due to increased utilization of these compounds in suggest its hypophosphorylation and suppression of protein an effort to offset the depletion of tricarboxylic acid cycle translation, consistent with general reduction of anabolism under intermediates. However, further studies are needed to confirm Icmt inhibition. It is important to note that suppression of this speculation. anabolism by Icmt inhibitor is not limited to PC3 cells; similar Cell anabolism, an integral feature of growth and proliferation, changes were observed in MDA-MB-231 cells (Supplementary depends on abundant supply of metabolites and high-energy Figure 5E), suggesting a general regulatory role of Icmt in cell molecules; therefore, we evaluated signatures of various macro- metabolism. molecule synthetic pathways in cells treated by Icmt inhibitor. Quantitative PCR analysis revealed that the expression of ACC, The impact on PC3 and MDA-MB-231 cancer cell viability by Icmt fatty acid synthase and 3-hydroxy-3-methylglutaryl-CoA reduc- inhibitor cysmethynil is mediated by its suppression of tase, important enzymes of fatty acid and cholesterol mitochondrial respiration synthesis,17,18 were significantly reduced upon cysmethynil As recent studies have demonstrated the potential of targeting treatment in PC3 cells (Figures 5c–e). Similar reduction of Icmt in inhibiting cancer cell growth, proliferation and survival,2,6,8 expression was also observed in MDA-MB-231 cells it was important to address the issue of whether Icmt-induced (Supplementary Figures 5B–D). ACC, fatty acid synthase and suppression of cellular proliferation and survival is mediated 3-hydroxy-3-methylglutaryl-CoA reductase expression are under through its effect on cellular respiration. We used three common transcriptional control, responding to the AMPK- approaches to address this question. catalyzed inhibitory phosphorylation.17 Therefore, AMPK is likely The first approach used mitochondrial DNA-null (ρ0) cells; an important signaling component between energy and without mitochondrial DNA, these cells are defective in oxidative

DMSO Cysm p-value 2.5 24h 48h 1.2 600 ± ± 2.0 Gly 0.20 0.10 0.12 0.09 0.241 0.9 Val 0.40 ± 0.08 0.32 ± 0.07 0.176 1.5 500 0.6 Leu/Ile 0.56 ± 0.16 0.43 ± 0.15 0.215 Expression 1.0 E xpression 0.3 Phe 0.33 ± 0.08 0.24 ± 0.07 0.131 0.5

400 FAS

± ± ACC1/2

cells Tyr 0.40 0.11 0.25 0.10 0.055

0.0 0.0 6 0 20 22.5 0 20 22.5 24h 48h 300 Ser 4.13 ± 0.77 3.12 ± 0.71 0.195 Cysm (μM) Glx 8.09 ± 1.99 4.96 ± 1.84 0.046

pmol/10 200 Pro 1.04 ± 0.23 0.81 ± 0.21 0.182 Asx 0.79 ± 0.14 0.27 ± 0.13 0.002 1.5 100 24 h 48 h ± ± Orn 0.02 0.01 0.014 0.01 0.265 1.2 Cysm 0 Cit 0.01 ± 0.01 0.01 ± 0.00 0.877 0.9 p-4EBP1

Arg 0.20 ± 0.06 0.14 ± 0.05 0.190 Expression 0.6 4EBP1 Ala 0.15 ± 0.02 0.12 ± 0.02 0.096 0.3 GAPDH

Met 0.12 ± 0.03 0.09 ± 0.03 0.185 HMGCR 0 His 0.09 ± 0.03 0.06 ± 0.02 0.138 24h 48h Figure 5. Icmt inhibitor treatment reduces important metabolites and anabolic activities in cancer cells. (a) Quantitation of organic acids in control (black bars) and cysmethynil-treated (gray bars) MDA-MB-231 cells. Organic acids were analyzed using isotope dilution techniques using Trace Ultra GC coupled to ISQ MS operating under Xcalibur 2.2 (Thermo Scientific).40 Error bars denote standard deviation of quadruplicate determinations; *Po0.05. (b) Quantitation of amino acids in control and cysmethynil-treated MDA-MB-231 cells. Amino acids were analyzed by flow injection electrospray ionization tandem mass spectrometry and quantified by isotope or pseudoisotope dilution using 35 previously developed methods. For both types of profiling (a and b), cells were treated with either DMSO or 20.0 μM cysmethynil for 24 h before being processed for the analysis. (c–f) Cysmethynil treatment reduces cellular anabolic activities. Quantitative PCR analysis of the expression levels of (c) ACC1 (black bar) and ACC2 (gray bar), (d) fatty acid synthase and (e) 3-hydroxy-3-methylglutaryl-CoA reductase in control and cysmethynil-treated PC3 cells. The cells were treated with DMSO, 20.0 or 22.5 μM cysmethynil, represented by black, gray and white bars, respectively in (d) and (e), for 24 or 48 h before qPCR analysis. 18S ribosome gene expression was used as normalization control for the quantification; error bars denote standard deviation. (f) Immunoblot analysis of 4EBP1 and phospho-4EBP1 in PC3 cells subjected to DMSO vehicle or increasing doses of cysmethynil (Cysm) at 17.5, 20.0 and 22.5 μM for the indicated durations.

© 2015 Macmillan Publishers Limited Oncogene (2015) 3296 – 3304 Icmt regulates mitochondrial respiration JT Teh et al 3302 phosphorylation.23–26 The hypothesis is that if cysmethynil viability from the cysmethynil treatment in PC3 cells, supporting reduces cell viability through inhibition of oxidative phosphoryla- the notion that this reduction in cell viability is through inhibition tion, then the ρ0 cells should be more resistant to the treatment of oxidative phosphorylation (Figure 6c). than the parental cells. We generated ρ0 cells from PC3 cells, using Last, we investigated the impact on cell viability of the the reported standard ethidium bromide protocol.27,28 Following concurrent treatment of cysmethynil and electron transport chain PCR confirmation for the lack of mitochondrial DNA in ρ0 complex I inhibitor rotenone and complex II inhibitor malonate. (Figure 6a), the two cell lines were subjected to Icmt inhibitor The anticipated outcome is that, if the biologic impact of treatment and viability determination. Indeed, ρ0 PC3 cell showed cysmethynil is similar to that of rotenone and malonate, the significant resistance to cysmethynil as compared with parental combination of these agents would generate little additional PC3 cells (Figure 6b), consistent with the notion that Icmt effect on viability. Consistent with the ρ0 and the rescue studies inhibition impacts on cancer cell viability via its inhibition of described above, rotenone and malonate were unable to exert mitochondrial function. further reduction of viability in cysmethynil-treated cells as In the second approach to determine the importance of compared with the cells treated with cysmethynil alone impairing respiration in the response to Icmt inhibition, we (Figure 6d). Taken together, the results from these three different attempted to rescue Icmt-treated cells with pyruvate and uridine, approaches strongly suggest that inhibition of cellular respiration the two supplements used to keep ρ0 cells alive and is a major cause of the suppression of cell proliferation and proliferating.28 Uridine rescues ρ0 cells because a critical reduction in survival that accompanies Icmt inhibition. pyrimidine biosynthetic enzyme, dihydrooratate dehydrogenase, requires mitochondrial electron transport for its normal redox DISCUSSION reaction.29,30 Pyruvate reduction to lactate is coupled to NADH oxidation to NAD+; NAD+ is essential for glycolysis, the sole It has been established by previous studies of ours and others that mean for energy production with dysfunctional oxidative Icmt inhibition impairs cell growth and proliferation, and therefore 7,8,23 phosphorylation.27,28 The hypothesis is that if the dysfunction of suppresses tumorigenesis. Efforts of our lab and others to cell respiration resulted from Icmt inhibition is the culprit of investigate the mechanism of these consequences of Icmt observed decrease in cell viability, the two supplements should be inhibition had focused on the usual suspects of growth factor able to rescue cysmethynil-treated cells. Indeed, the presence of related signaling, for example, the phosphatidylinositol 3-kinase- uridine and pyruvate almost completely rescued the reduction of AKT and mitogen-activated protein kinase axis, classic pathways of tumorigenesis.24 The current findings extend the repertoire of critical cellular processes impacted by Icmt function to include 110 oxidative phosphorylation. More importantly, the study provided strong evidence to suggest that the effect on oxidative PC3 Mitochondrial DNA (%) 80 phosphorylation account for the impact of Icmt inhibition on Parental cells 101.04 ± 17.17 cancer cell growth and survival. Targeting mitochondrial respiratory function can be explored in ρ0 cells 0.0017 ± 0.0003 50 cancer therapy. The maximum ability of oxidative phosphorylation p-value 0.0005 Cell Viability (%) 20 of a cell, termed mitochondrial respiratory capacity, is a critical Cysm (μM) determinant of cell survival under stressful conditions. Such is the environment of many tumors, wherein the cancer cells are subjected to nutrient limitation and hypoxia, which make them 110 120 more vulnerable to Icmt inhibition. Indeed, mitochondrial respiration can be very important for tumor cell survival.27,29 This particular vulnerability of cancer cells has already been exploited; 80 80 several drugs acting on mitochondrial electron transport chain, which inhibit complex I and II, respectively, have been reported to 31,32 50 40 have anticancer effects. The energy depletion and reduced anabolism are at odds with the fundamental characteristics of Cell Viability (%) Cell Viability (%) 20 cancer cell, which requires energy and biosynthesis for uncon- 0 trolled growth and proliferation. Indeed, the marked reduction of Cysm (μM) DMSO Cysm NTPs and small-molecule building blocks and inhibition of fatty Figure 6. Cysmethynil decreases cancer cell viability through its acid, cholesterol and protein synthesis signaling support the role suppression of mitochondrial respiration. (a) qPCR confirmation of of Icmt inhibition in devastating cancer cell metabolic need. We the absence of mitochondrial DNA in selected ρ0 PC3 cells. (b) MTS ρ postulate that some cancer cells may be more vulnerable to Icmt analysis of cell viability of 0 PC3 cells (gray line) and wild-type PC3 inhibition owing to the differences in mitochondria respiration cells (black line) after treatment with cysmethynil for 24 h. The gray and metabolism of these cells. More importantly, further triangle under horizontal axis represents increasing cysmethynil fi fi concentrations of 20–24 μM. Data represent the mean and standard speci city and ef cacy of targeting Icmt as means for cancer deviation of triplicate determinations. (c) Analysis of viability of therapy will depend on the identification of the downstream Icmt PC3 cells supplemented with (gray line) or without (black line) substrate(s), which regulate the activity of mitochondrial com- uridine (50 μg/ml) and pyruvate (100 μg/ml) with cysmethynil plexes. Therefore, inhibition of Icmt can provide an anticancer treatment for 24 h. The cysmethynil concentrations are the same strategy by targeting cancer cell energy metabolism, with added as in (b). Data represent the mean and standard deviation of specificity on cancer cells bearing oncogenic prenylated protein(s). triplicate determinations. (d) Analysis of PC3 cell viability following combined treatment of cysmethynil with rotenone and malonate for 24 h. The cells were split into two treatment groups: 20 μM MATERIALS AND METHODS cysmethynil and DMSO (control). Both groups of cells were subjected to three treatment conditions, DMSO (black bar), 10 nM Antibodies, cell lines, cell culture, drug treatment and cell viability rotenone+10 μM malonate (gray bar) or 1 μM rotenone+1 mM assays malonate (white bar). Data represent the mean and standard Antibodies for phospho-AMPK, phospho-ACC, GAPDH and ATG5 were from deviation of triplicate determinations. Experiments in (b–d) have Cell Signaling Technology (Danver, MA, USA); LC3-specific antibody was been performed three times with similar results. from Abgent (San Diego, CA, USA); mitochondrial complex antibodies for

NDUFS3, SDHA, UQCRC2, COXIV and ATP5A were from Abcam antimycin (1 μM). Complex IV respiration is measured by oxygen consump- (Cambridge, UK). tion after addition of ascorbate (10 mM) and TMPD (0.2 mM) and suppressed PC3 and MDA-MB-231 cell lines were obtained from American Type by KCN (700 μM). All reagents were from Sigma-Aldrich. Culture Collection (Rockville, MD, USA); wild-type and Icmt − / − murine embryonic fibroblast were gifts from Dr Martin Bergo, Sahlgrenska University (Göteborg, Sweden). Cells were cultured in Dulbecco’s modified In vitro assays for mitochondrial complex I, II, IV and V activities Eagle’s medium (Sigma-Aldrich, St Louis, MO, USA) supplemented with In vitro mitochondrial complex I, II, IV and V activities were assessed using 10% fetal bovine serum (Hyclone, Logan, UT, USA), 50 U/ml penicillin and total cell lysates and were measured using kits from Abcam according to 50 μg/ml streptomycin (Gibco, Gaithersburg, MD, USA), at 37 1C with 5% the manufacturer’s instructions. Briefly, the cells were harvested and CO2. Cysmethynil treatment of cells performed with concentrations homogenized using a standard protocol, and samples were adjusted to the ranging from 17.5 to 25 μM according to the sensitivity of different cell concentration of 5.5 mg/ml using PBS. Subsequently, 1/10 of the volume of ® lines. All cell viability studies were carried out using CellTiter 96 Aqueous the detergent solution (Abcam) was added to each sample and the One Solution cell proliferation assay (Promega, Madison, WI, USA); specific mixture was incubated for 30 min on ice to extract the mitochondrial condition for each experiment is described in the figure legends. proteins. The samples were centrifuged for 20 min at 21 000 g and the supernatant was collected. The samples were then diluted to optimal fi ρ0 concentration using solution buffer provided in the kit, before being added Generation of mitochondrial DNA-de cient cell line ( cells) to microplates coated with antibodies against each electron transport ρ0 PC3 cell line was generated according to the protocol of King and chain complex to capture the mitochondrial complex. The complex 28 fl Attardi. Brie y, cells were cultured as described above and selected using activities were determined calorimetrically using Tecan Infinite200 Micro- 250 ng/ml ethidium bromide (Sigma-Aldrich), supplemented with 1 mM plate Reader (Männedorf, Switzerland). sodium pyruvate (Sigma-Aldrich) and 50 mg/ml uridine (Sigma-Aldrich) for 4 weeks. The lack of mitochondrial DNA in ρ0 cells was confirmed using reverse transcription–PCR. Cysmethynil treatment and cell viability assay Metabolomic analysis was performed as described above. Supplementation of pyruvate and Nucleotides were measured by a previously reported LC-MS/MS method.34 uridine was halted during cysmethynil treatment. Amino acids were analyzed by flow injection electrospray ionization tandem mass spectrometry and quantified by isotope or pseudoisotope 35–39 Transfection dilution using previously developed methods. Organic acids were analyzed by capillary gas chromatography/mass spectrometry using Transfections of siRNA were performed with Lipofectamine TM 2000 and ’ isotope dilution techniques using Trace Ultra GC coupled to ISQ MS OptiMEM (Invitrogen, North Andover, MA, USA) as per the manufacturer s operating under Xcalibur 2.2 (Thermo Scientific, Waltham, MA, USA).40 protocol. siRNA targeting Atg5, Atg7 and Icmt were from Invitrogen.

Real-time PCR and mitochondrial DNA quantification CONFLICT OF INTEREST RNA extraction, cDNA preparation and quantitative real-time PCR were The authors declare no conflict of interest. carried out using the same protocol as reported in our past studies.33 For mitochondrial DNA analysis, DNA was extracted using DNeasy Blood and Tissue Kit (Qiagen, Hilden, Germany) as per the manufacturer’s manual; ACKNOWLEDGEMENTS quantitative real-time PCR was then carried out using the extracted DNA as These experiments were supported by grant from National Medical Research Council template. Sequences of primers probing human mitochondrial DNA and (NMRC) and Duke-NUS institutional funding. Wild-type and Icmt-null MEFs are from human genomic DNA were described as below. Mitochondrial DNA: Dr M Bergo. forward primer, 5′-TCT GTA TGT CTC CAT CTA TTG ATG AGG GTC T-3′ and reverse primer, 5′-ATG TTG TCA AAA CTA GTT AAT TGG AAG TTA ACG GT-3′; genomic DNA: forward primer, 5′-CCGCGCCCCCGGTTTCTATA-3′ and reverse REFERENCES primer, 5′-CCC GTC TTC ACC TGG CGA CG-3′. Quantitative analysis was 9 – carried out by normalizing the amount of mitochondrial DNA to that of 1 Ashby MN. CaaX converting enzymes. Curr Opin Lipidol 1998; :99 102. genomic DNA. 2 Winter-Vann AM, Casey PJ. Post-prenylation-processing enzymes as new targets in oncogenesis. Nat Rev Cancer 2005; 5: 405–412. 3 Seabra MC. Membrane association and targeting of prenylated Ras-like GTPases. Mito stress and glycolytic stress test assays Cell Signal 1998; 10: 167–172. Cells were seeded, cultured and treated by drugs in XF24 cell culture plates 4 Bergo MO, Leung GK, Ambroziak P, Otto JC, Casey PJ, Young SG. Targeted as detailed in the respective figure legends. One hour before performing inactivation of the isoprenylcysteine carboxyl methyltransferase gene causes the OCR or ECAR assays, media were replaced by XF assay medium mislocalization of K-Ras in mammalian cells. J Biol Chem 2000; 275: 17605–17610. (Seahorse Bioscience, North Billerica, MA, USA) and incubated at 37 1Cina 5 Cushman I, Casey PJ. RHO methylation matters: a role for isoprenylcysteine 5 CO2-free environment. The subsequent Mito stress assay for OCR and carboxylmethyltransferase in cell migration and adhesion. Cell Adhes Migr 2011; : glycolytic stress test assay (ECAR) were performed as per the XF24 analyzer 11–15. standard protocol (Seahorse Bioscience). Oligomycin, FCCP, rotenone and 6 Wahlstrom AM, Cutts BA, Liu M, Lindskog A, Karlsson C, Sjogren AK et al. antimycin A are from Seahorse Bioscience; glucose, oligomycin and 2- Inactivating Icmt ameliorates K-RAS-induced myeloproliferative disease. Blood 112 – deoxy-D-glucose are purchased from Sigma-Aldrich. Both OCR and ECAR 2008; : 1357 1365. were measured by the XF24 analyzer (Seahorse Bioscience). 7 Bergo MO, Gavino BJ, Hong C, Beigneux AP, McMahon M, Casey PJ et al. Inactivation of Icmt inhibits transformation by oncogenic K-Ras and B-Raf. J Clin Invest 2004; 113: 539–550. In vivo oximetry 8 Wang M, Tan W, Zhou J, Leow J, Go M, Lee HS et al. A small molecule inhibitor of One or two million cells were assayed in 2 ml MiR05 medium (0.5 mM EGTA, isoprenylcysteine carboxymethyltransferase induces autophagic cell death in PC3 3mM MgCl2,60mM K-lactobionate, 20 mM taurine, 10 mM KH2PO4,20mM prostate cancer cells. J Biol Chem 2008; 283: 18678–18684. HEPES, 110 mM sucrose, 1 g/l bovine serum albumin) using the Oroboros 9 Wang M, Hossain MS, Tan W, Coolman B, Zhou J, Liu S et al. Inhibition of Oxygraph-2k respirometer (Oroboros Instruments, Innsbruck, Austria). isoprenylcysteine carboxylmethyltransferase induces autophagic-dependent Respiratory oxygen consumption was assessed in real-time as pmol of O2 apoptosis and impairs tumor growth. Oncogene 2010; 29: 4959–4970. per s per million cells. Digitonin (10 μg per million cells for PC3, 5 μgper 10 Mihaylova MM, Shaw RJ. The AMPK signalling pathway coordinates cell growth, million cells for MDA-MB-231 and 17.5 μg per million cells for wild-type and autophagy and metabolism. Nat Cell Biol 2011; 13: 1016–1023. Icmt-null MEF) were first added to permeabilize the cells. Subsequently, 11 Kim J, Kundu M, Viollet B, Guan KL. AMPK and mTOR regulate autophagy through pyruvate (5 mM), malate (2 mM)andADP(1mM) were added and the direct phosphorylation of Ulk1. Nat Cell Biol 2011; 13:132–141. respiration was measured. The reaction is then inhibited by rotenone (1 μM). 12 Krebs HA, Hems R. Phosphate-transfer reactions of adenosine and inosine Succinate (5 mM) oxidation was measured and then inhibited by malonate nucleotides. Biochem J 1955; 61:435–441. (10 mM)orantimycin(1μM). Complex III respiration is measured by oxygen 13 Berg P, Joklik WK. Transphosphorylation between nucleoside polyphosphates. consumptions after the addition of duroquinol (0.6 mM) and suppressed by Nature 1953; 172: 1008–1009.

© 2015 Macmillan Publishers Limited Oncogene (2015) 3296 – 3304 Icmt regulates mitochondrial respiration JT Teh et al 3304 14 Mizushima N, Noda T, Yoshimori T, Tanaka Y, Ishii T, George MD et al. A protein 28 King MP, Attardi G. Isolation of human cell lines lacking mitochondrial DNA. conjugation system essential for autophagy. Nature 1998; 395: 395–398. Methods Enzymol 1996; 264:304–313. 15 Suzuki K, Kirisako T, Kamada Y, Mizushima N, Noda T, Ohsumi Y. The 29 Vander Heiden MG, Cantley LC, Thompson CB. Understanding the Warburg effect: pre-autophagosomal structure organized by concerted functions of APG genes is the metabolic requirements of cell proliferation. Science 2009; 324: 1029–1033. essential for autophagosome formation. EMBO J 2001; 20: 5971–5981. 30 Gregoire M, Morais R, Quilliam MA, Gravel D. On auxotrophy for pyrimidines of 16 Barrientos A. In vivo and in organello assessment of OXPHOS activities. Methods respiration-deficient chick embryo cells. Eur J Biochem/FEBS 1984; 142:49–55. 2002; 26: 307–316. 31 Neuzil J, Dong LF, Ramanathapuram L, Hahn T, Chladova M, Wang XF et al. 17 Li Y, Xu S, Mihaylova MM, Zheng B, Hou X, Jiang B et al. AMPK phosphorylates and Vitamin E analogues as a novel group of mitocans: anti-cancer agents that act by inhibits SREBP activity to attenuate hepatic steatosis and atherosclerosis in targeting mitochondria. Mol Aspects Med 2007; 28(5-6): 607–645. diet-induced insulin-resistant mice. Cell Metab 2011; 13:376–388. 32 Pathania D, Millard M, Neamati N. Opportunities in discovery and delivery of 18 Inoki K, Zhu T, Guan KL. TSC2 mediates cellular energy response to control cell anticancer drugs targeting mitochondria and cancer cell metabolism. Adv Drug growth and survival. Cell 2003; 115:577–590. Deliv Rev 2009; 61: 1250–1275. 19 Carling D, Zammit VA, Hardie DG. A common bicyclic protein kinase cascade 33 Zhu WL, Hossain MS, Guo DY, Liu S, Tong H, Khakpoor A et al. A role for Rac3 inactivates the regulatory enzymes of fatty acid and cholesterol biosynthesis. FEBS GTPase in the regulation of autophagy. J Biol Chem 2011; 286: 35291–35298. Lett 1987; 223:217–222. 34 Cordell RL, Hill SJ, Ortori CA, Barrett DA. Quantitative profiling of nucleotides and 20 Sato R, Goldstein JL, Brown MS. Replacement of serine-871 of hamster 3-hydroxy- related phosphate-containing metabolites in cultured mammalian cells by liquid 3-methylglutaryl-CoA reductase prevents phosphorylation by AMP-activated chromatography tandem electrospray mass spectrometry. J Chromatogr B 2008; kinase and blocks inhibition of sterol synthesis induced by ATP depletion. Proc 871:115–124. Natl Acad Sci USA 1993; 90: 9261–9265. 35 Wu JY, Kao HJ, Li SC, Stevens R, Hillman S, Millington D et al. ENU mutagenesis 21 Hara K, Yonezawa K, Kozlowski MT, Sugimoto T, Andrabi K, Weng QP et al. identifies mice with mitochondrial branched-chain aminotransferase deficiency Regulation of eIF-4E BP1 phosphorylation by mTOR. J Biol Chem 1997; 272: resembling human maple syrup urine disease. J Clin Invest 2004; 113:434–440. 26457–26463. 36 An J, Muoio DM, Shiota M, Fujimoto Y, Cline GW, Shulman GI et al. Hepatic 22 Gingras AC, Gygi SP, Raught B, Polakiewicz RD, Abraham RT, Hoekstra MF et al. expression of malonyl-CoA decarboxylase reverses muscle, liver and whole- Regulation of 4E-BP1 phosphorylation: a novel two-step mechanism. Genes Dev animal insulin resistance. Nat Med 2004; 10:268–274. 1999; 13: 1422–1437. 37 Millington DS, Terada N, Chace DH, Chen YT, Ding JH, Kodo N et al. The role of 23 Winter-Vann AM, Baron RA, Wong W, dela Cruz J, York JD, Gooden DM et al. tandem mass spectrometry in the diagnosis of fatty acid oxidation disorders. A small-molecule inhibitor of isoprenylcysteine carboxyl methyltransferase with Progr Clin Biol Res 1992; 375:339–354. antitumor activity in cancer cells. Proc Natl Acad Sci USA 2005; 102: 4336–4341. 38 Millington DS, Stevens RD. Acylcarnitines: analysis in plasma and whole blood 24 Pylayeva-Gupta Y, Grabocka E, Bar-Sagi D. RAS oncogenes: weaving a using tandem mass spectrometry. Methods Mol Biol 2011; 708:55–72. tumorigenic web. Nat Rev Cancer 2011; 11:761–774. 39 Chace DH, Millington DS, Terada N, Kahler SG, Roe CR, Hofman LF. Rapid diagnosis 25 King MP, Attardi G. Human cells lacking mtDNA: repopulation with exogenous of phenylketonuria by quantitative analysis for phenylalanine and tyrosine in mitochondria by complementation. Science 1989; 246: 500–503. neonatal blood spots by tandem mass spectrometry. Clin Chem 1993; 39:66–71. 26 Chandel NS, Schumacker PT. Cells depleted of mitochondrial DNA (rho0) yield 40 Jensen MV, Joseph JW, Ilkayeva O, Burgess S, Lu D, Ronnebaum SM et al. Com- insight into physiological mechanisms. FEBS Lett 1999; 454:173–176. pensatory responses to pyruvate carboxylase suppression in islet beta-cells. 27 Potter VR. The biochemical approach to the cancer problem. Fed Proc 1958; 17: Preservation of glucose-stimulated insulin secretion. J Biol Chem 2006; 281: 691–697. 22342–22351.

Supplementary Information accompanies this paper on the Oncogene website (http://www.nature.com/onc)