Oncogenic K-Ras Expression Is Associated with Derangement of the Camp/PKA Pathway and Forskolin-Reversible Alterations of Mitochondrial Dynamics and Respiration

ORIGINAL ARTICLE Oncogenic K-ras expression is associated with derangement of the cAMP/PKA pathway and forskolin-reversible alterations of mitochondrial dynamics and respiration

R Palorini1, D De Rasmo2,3, M Gaviraghi1, L Sala Danna1, A Signorile2, C Cirulli1, F Chiaradonna1, L Alberghina1 and S Papa2,3

The Warburg effect in cancer cells has been proposed to involve several mechanisms, including adaptation to hypoxia, oncogenes activation or loss of oncosuppressors and impaired mitochondrial function. In previous papers, it has been shown that K-ras transformed mouse cells are much more sensitive as compared with normal cells to glucose withdrawal (undergoing apoptosis) and present a high glycolytic rate and a strong reduction of mitochondrial complex I. Recent observations suggest that transformed cells have a derangement in the cyclic adenosine monophosphate/cAMP-dependent protein kinase (cAMP/PKA) pathway, which is known to regulate several mitochondrial functions. Herein, the derangement of the cAMP/PKA pathway and its impact on transformation-linked changes of mitochondrial functions is investigated. Exogenous stimulation of PKA activity, achieved by forskolin treatment, protected K-ras-transformed cells from apoptosis induced by glucose deprivation, enhanced complex I activity, intracellular adenosine triphosphate (ATP) levels, mitochondrial fusion and decreased intracellular reactive oxygen species (ROS) levels. Several of these effects were almost completely prevented by inhibiting the PKA activity. Short-time treatment with compounds favoring mitochondrial fusion strongly decreased the cellular ROS levels especially in transformed cells. These findings support the notion that glucose shortage-induced apoptosis, specific of K-ras-transformed cells, is associated to a derangement of PKA signaling that leads to mitochondrial complex I decrease, reduction of ATP formation, prevalence of mitochondrial fission over fusion, and thereby opening new approaches for development of anticancer drugs.

Oncogene (2013) 32, 352--362; doi:10.1038/onc.2012.50; published online 12 March 2012 Keywords: cancer cell metabolism; oncogenic K-ras; PKA; mitochondrial morphology; mitochondrial activity; glucose deprivation

INTRODUCTION a depression of genes encoding for components of the cyclic Tumors are generally characterized by a high rate of anaerobic adenosine monophosphate/cAMP-dependent protein kinase 10,26 --31 and aerobic glycolysis and a reduced rate of respiration.1--6 This (cAMP/PKA) signaling pathway has also been reported. switch of energy metabolism is known as the ‘Warburg effect’ and The cAMP/PKA pathway is known to regulate cellular energy it has been reported to be linked to mitochondrial dysfunctions metabolism, exerting a stimulatory effect on glucose transport and 32 --34 that lead cells to depend on glycolysis to generate adenosine utilization, glycolysis, glycogen breakdown and glucose 35 triphosphate (ATP).7--9 The genetic and phenotypic features oxidation. The cAMP/PKA pathway has recently been found to 36 --39 40,41 involved in this general derangement of cellular energy metabo- regulate also mitochondrial respiration, dynamics and 42,43 lism in tumors are not yet fully understood.3,5 However, there are apoptosis. Proteins belonging to the cAMP/PKA pathway are 44 several evidences that alterations in signaling pathways, which associated with various cellular compartments and structures 45 regulate glucose uptake and utilization and mitochondrial including the outer membrane and the inner compartment of 46 --49 functions, may contribute to the Warburg effect.10 --13 In addition, mitochondria. The cytosolic pool of cAMP, produced by the cancer cells mitochondria have been reported to be associated plasma membrane adenylyl cyclase, enhances the functional with mitochondrial DNA mutations,14,15 altered expression capacity of complex I (NADH-ubiquinone oxidoreductase) of the of mitochondrial enzymes,16 reduced levels and activity of respiratory chain and reduces reactive oxygen species (ROS) 37,50 mitochondrial OXPHOS complexes17 --19 and deregulation of production through activation of PKA. mitochondrial fusion and ﬁssion.20 In recent years, an in vitro model of cellular transformation In various human tumors (up to 35%), the oncogenic ras gene has been developed. It is represented by two NIH3T3 derived has been found mutated21,22 and these mutations have a critical stable cell lines, NIH3T3 cells expressing an oncogenic K-ras role in the onset of different malignant phenotypes. The ras genes (transformed cell line) and NIH-ras cells expressing a GEF-DN, a encode for guanidine nucleotide binding proteins that mediate protein able to downregulate ras activation and to phenotipically 13,51 cellular signal transduction pathways controlling cell growth and revert transformed cells (reverted cell line). K-ras cell line has differentiation.21,22 Although active form of the ras family been extensively characterized and found to have several members, such as K-ras, have been related to enhancement of metabolic alterations.13,52,53 It has been shown that the enhanced intracellular level of cyclic adenosine monophosphate (cAMP), proliferation potential of K-ras-transformed cells requires high through activation of the plasma membrane adenylyl cyclase,23 --25 initial glucose and glutamine concentrations in the medium.

1Department of Biotechnology and Biosciences, University of Milano-Bicocca, Milan, Italy; 2Department of Medical Biochemistry, Biology and Physics (DIBIFIM), University of Bari, Bari, Italy and 3Institute of Biomembranes and Bioenergetics (IBBE), Consiglio Nazionale delle Ricerche, Bari, Italy. Correspondence: Dr F Chiaradonna or Professor L Alberghina, Department of Biotechnology and Biosciences, University of Milano-Bicocca, Piazza della Scienza 2, 20126 Milan (MI), Italy. E-mails: [email protected] or [email protected] Received 30 July 2011; revised 27 December 2011; accepted 19 January 2012; published online 12 March 2012 PKA regulation of K-ras cancer cell mitochondria R Palorini et al 353 The selective growth advantage of transformed cells is lost in normal cells that was significantly smaller at 1 mM glucose. Thus, sub-optimal glucose or glutamine growth conditions. Glucose and these data show that in transformed cells the mitochondrial glutamine-dependent growth correlates with an altered metabolic oxidative phosphorylation contributes only marginally to cellular pattern namely increased glucose utilization and lactate produc- ATP production, which is essentially contributed by glycolysis.13 tion, increased utilization of glutamine through the tricarboxylic acid cycle, increased expression of glycolytic genes, depressed cAMP/PKA pathway alterations in transformed fibroblasts expression of mitochondrial genes, altered mitochondrial morphology and reduced ability of ATP production.13,52,54 Trans- Direct determination of basal endogenous levels of cAMP along a formed cells produce large amount of ROS associated with time course of 72 h in cells grown at 25 and 1 mM glucose decreased activity of mitochondrial complex I55 and increased cell (Figures 2a and b) showed a time-dependent increase in cAMP death.13 Remarkably, several metabolic phenotypes induced by levels in both cell lines. In transformed cells, at both 48 and 72 h, oncogenic K-ras expression are almost completely reverted by the cAMP levels were higher than in normal cells (Figures 2a and b). GEF-DN expression52 --54 suggesting a main role of oncogenic ras The addition of FSK (an activator of the adenylyl cyclases) signaling in the induction and maintenance of metabolic increased, as expected, the level of cAMP, but such an increase alterations. was much larger in transformed cells as compared with normal This study was aimed at examining alterations of the cAMP/PKA ones (Figures 2a and b). pathway in K-ras-transformed murine NIH3T3 fibroblasts and Kinetic analysis of the PKA activity in cell extracts showed in human breast cancer MDA-MB-231 cells and their impact on cell transformed cells a constitutive (time 0 of the growth curve) growth and energy metabolism. depressed activity of the enzyme as compared with normal cells, The results show that in both cell lines there is an alteration of which slightly increased in 48 h growth and was greatly enhanced the cAMP/PKA system, characterized by enhanced excitability of the by FSK addition to the cultivation medium (Figure 2d). In normal plasma membrane adenylyl cyclase but a reduced PKA activity. This cells, the PKA activity decreased significantly during cell growth alteration of the cAMP/PKA pathway in oncogenic K-ras cells was and was further depressed by FSK (Figure 2c). Direct immuno- associated with a shift of mitochondrial dynamics to a more chemical analysis revealed an increase in the cellular content of fragmented structure, increased apoptosis under glucose depriva- the catalytic subunit of PKA in 48 h of growth of normal and tion, deficiency of complex I activity and ATP production and transformed cells, which in both type of cells was prevented by increase in intracellular ROS levels. Forskolin (FSK), a known FSK addition (Figures 2e and f). These findings indicate that the activator of adenylyl cyclase, leading to an increase in the cytosolic changes in the functional capacity of PKA, rather than being due level of cAMP and hence to PKA activation, induced a reversion of to changes in the content of the enzyme, reflect altered PKA all these changes and promoted survival of oncogenic K-ras cells. responsiveness to cellular activator(s)/inhibitor(s). In order to evaluate the PKA activation state as a function of glucose availability and of oncogenic ras mutation, the extent of RESULTS cAMP responsive element binding protein (CREB) phosphorylation K-ras transformation induces marked glucose dependence of cell was monitored in normal and transformed cells at both 25 and growth 1mM glucose. In normal cells (Figure 3a), as well in reverted cells NIH3T3 cells are a genetically well-characterized immortalized cell (Supplementary Figures S2A and S2B), phosphorylation of CREB line that has long been established as a model of ‘normal’ cells for significantly increased with the cultivation time, in particular at the study of cell transformation, as these cells undergo contact 25 mM glucose (Figure 3a). On the contrary, in transformed cells no inhibition, exhibit no growth in soft agar and do not form tumors significant time-dependent increase in CREB phosphorylation in immunocompromised mice,56 in contrast to isogenic trans- occurred either at 25 or 1 mM glucose (Figure 3b). FSK treatment formed lines as K-ras NIH3T3 cells.51 Besides, as previously resulted in a transient increase in CREB phosphorylation, which described, oncogenic ras proteins expression correlates with the was larger and more persistent in transformed cells as compared appearance of several characteristic metabolic alterations of with normal ones (Figures 3c and d). Altogether, these findings 13,21,57 --59 show an altered regulation of the cAMP/PKA pathway in cancer cells. Given that both Ras and cAMP/PKA pathways are able to transformed cells. control cellular metabolism and proliferation,60 we decided to test the relationship between the cAMP/PKA system and the Effects of FSK on cell growth, apoptosis and mitochondrial proliferation ability of NIH3T3 and K-ras NIH3T3 cells on alteration dynamics of glucose availability. To gain an insight into the impact of the above-detected Asynchronous NIH3T3 (normal) and K-ras NIH3T3 (transformed) alterations of the cAMP/PKA pathway on mitochondrial dynamics cell lines were cultured in normal growth medium (25 mM and functions in K-ras-transformed cells, the effect of adenylyl glucose), in a low glucose medium (1 mM glucose), in glucose- cyclase activation by FSK was tested on cell growth, apoptosis, free medium supplemented with 25 mM galactose, an obligate mitochondrial structure and functions of normal and transformed mitochondrial oxidative substrate, or in 5 mM fructose. The growth fibroblasts. of normal cells declined after 72 h in both 25 and 1 mM glucose Proliferation of transformed cells grown at 1 mM glucose began and at 48 h in galactose (Figures 1a and c) and fructose to slow down at 48 h and under these conditions they started to (Supplementary Figure S1A). Transformed cells maintained a die and undergo apoptosis more significantly and faster than vigorous growth for at least 96 h in 25 mM glucose but lost their normal cells (Figure 4). Addition of FSK to the culture medium proliferative ability in 1 mM glucose (Figure 1b), in 25 mM galactose prevented the glucose-dependent death of transformed cells (Figures 1b and c) and in 5 mM fructose (Supplementary Figure (Figures 4b, d and f). The death-preventing effect of FSK was S1B). Addition to the culture medium of oligomycin (OM) suppressed by the addition of H89, a specific inhibitor of PKA62 (Figure 1d), a specific inhibitor of the mitochondrial FoF1 ATP (Supplementary Figure S3B). In normal (Figures 4a, c and e) and synthase,61 in normal cells reduced the ATP levels by around 30% reverted cells (Supplementary Figure S2C) an inhibiting effect of both at 25 and 1 mM glucose. Thus, this fraction represents the FSK on cell growth was observed. contribution of mitochondrial oxidative phosphorylation to ATP In order to investigate whether the survival of FSK-treated cells production under the prevailing cultivation conditions. In could be consequence of a different ability of transformed cells to transformed cells, cultivated in 25 mM glucose, OM addition transport and utilize glucose, glucose utilization as well as lactate resulted in a smaller reduction of the ATP levels as compared with production of both cell lines grown in 1 mM glucose condition

& 2013 Macmillan Publishers Limited Oncogene (2013) 352 --362 PKA regulation of K-ras cancer cell mitochondria R Palorini et al 354

Figure 1. K-ras-transformed cells are strongly sensitive to glucose availability. Proliferation curves of normal (N) (a) and transformed (T) (b) cells cultured at 25 mM glucose (Glc), at 1 mM Glc or at 25 mM galactose (Gal) were determined counting cells at indicated time points. Note the different scale for the two cell lines. (c) Percentage of proliferation reduction in 25 mM Gal as compared with 25 mM Glc at each time point for N and T cells. (d) Total intracellular ATP levels were measured in N and T cells grown at 25 and 1 mM Glc, treated or not with 5 mM OM at 24 h of culture. Values are relative to the sample 25 mM Glc. All data represent the average of at least three independent experiments (±s.d.).

cd cd

Figure 2. FSK treatment largely increases intracellular cAMP levels Figure 3. FSK treatment induces CREB phosphorylation especially in and PKA activity in transformed cells as compared with normal ones. transformed cells. (a, b) For basal expression analysis of total CREB (a, b) The cAMP levels were measured at indicated time points in (Creb) and phospho-CREB Ser133 (P-Creb), normal (a)andtransformed normal (a) and transformed (b) cells grown at 25 and 1 mM glucose (b) cells were collected at indicated time points and total cellular ex- (Glc), untreated and þ FSK. (c, d) PKA activity was measured as tracts were subjected to sodium dodecyl sulfate--polyacrylamide gel described in Materials and methods at indicated time points in electrophoresis followed by western blot analysis with specific anti- normal (c) and transformed (d) cells grown at 1 mM Glc (untreated bodies. Relative quantitative measurement (a and b,bottomhisto- and þ FSK). (e, f) For expression analysis of catalytic subunit of PKA grams) of CREB phosphorylation, in both cell lines grown at 25 or 1 mM (cat), normal (e) and transformed (f) cells, grown at 1 mM Glc glucose (Glc), was performed by densitometric analysis of western blot (untreated and þ FSK), were collected at indicated time points and films. The values obtained for P-CREB were normalized to the corre- total cellular extracts were subjected to sodium dodecyl sulfate-- sponding total CREB values, plotted as fold change from the sample 0 h polyacrylamide gel electrophoresis followed by western blot ana- (0h ¼ 1) and indicated as relative densitometric units (DUs). (c, d) lysis with a specific antibody. Relative quantitative measurement Western blot analysis of CREB phosphorylation (upper panels) and re- (e and f histograms) was performed by densitometric analysis of lative quantitative measurement (lower panels) were also performed in western blot films and normalized to the corresponding b-actin normal (c)andtransformed(d)cellsgrownat1mM glucose on 24 h of values. All data represent the mean ±s.e.m. of three independent FSK treatment. Values were normalized as described above. Data re- determinations. present the mean±s.e.m. of three independent determinations.

Oncogene (2013) 352 --362 & 2013 Macmillan Publishers Limited PKA regulation of K-ras cancer cell mitochondria R Palorini et al 355

Figure 4. FSK treatment increases survival of K-ras-transformed cells on glucose deprivation. (a, b) Proliferation curves of normal (a) and transformed (b) cells grown at 1 mM glucose (Glc), treated or not with FSK, were determined counting cells at indicated time points. (c, d) Phase contrast microscopy images were collected for normal (c) and transformed (d) cells at 96 h of cell culture. (e, f) Analysis of apoptosis for normal (e) and transformed (f) cells was evaluated by annexin-V/propidium iodide (PI) staining and calculating the percentage of annexin-V- positive cells. Data represent the average of at least three independent experiments (±s.d.). were assayed. As shown in Supplementary Figures S4A and B, the shown in Figures 5b and c, right panels, H89 strongly prevented amounts of consumed glucose and of secreted lactate, as the FSK-dependent formation of more interconnected and measured in the culture medium, were substantially identical intermediate mitochondria in transformed cells as compared with between untreated and FSK-treated cell lines. These findings normal ones. were confirmed also by measuring both glucose and lactate on per cell basis (data not shown) and by measuring the rate of glucose transport into the cells, using the fluorescent derivative Complex I activity, ROS production and ATP formation: effect of FSK of glucose 2-N-[(7-nitrobenz-2-oxa-1,3-diazol-4-yl)amino]-2-deoxy- In Figure 6, the results of an analysis of the functional capacity of D-glucose (2-NBDG) as indicator of glucose uptake (Supplementary complex I and IV of the respiratory chain, ATP and ROS production Figures S4C and D). in normal and transformed cells are presented at 72 h of growth in In stable clones of normal and transformed cell lines expressing 1mM glucose (48 h of FSK treatment). In transformed cells, the a mitochondrially targeted EYFP-Mito protein, the effect of FSK specific activity of complex I (NADH-ubiquinone oxidoreductase), treatment on mitochondrial morphology in 1 mM glucose culture depressed as compared with normal cells, was stimulated by FSK condition was investigated by fluorescent microscopy (Figure 5). treatment (Figure 6b), but it was unaffected in normal cells Three types of mitochondrial morphologies have been classified (Figure 6a). No changes were observed in the specific activity of as fragmented, intermediate and networked (Figure 5a). At an complex IV (cytochrome c oxidase) (Figure 6d). Transformed cells early cultivation time, normal cells displayed both fragmented and as compared with normal ones showed a slight but significant intermediate mitochondria (Figure 5b, left panel). At later increase in the respiratory activity with NAD-linked substrates, in cultivation time (72--96 h), fragmented mitochondria completely particular in the uncoupled state, on FSK treatment (data not disappeared and the vast majority of cells showed a networked shown). No effect of FSK was observed on the respiratory activity morphology (Figure 5b, left panel). In normal cells at all time with succinate as respiratory substrate (data not shown). The level points, morphology was not affected by FSK (Figure 5b, middle of cellular ATP was increased by the presence of FSK in both cell panel). On the contrary, transformed cells showed more frag- lines (Figures 6e and f). This increase was larger in transformed mented mitochondria and intermediate mitochondria along all cells (Figures 6e and f) and was almost completely prevented by the time course of analysis (Figure 5c, left panel) and were H89, especially in transformed cells as compared with normal cells strongly sensitive to FSK, which induced a marked decrease in (Figures 6e and f). Notably, the FSK-dependent effects on complex fragmented mitochondria with a parallel increase in intermediate I activity and ATP levels were also associated with a significant and networked structures (Figure 5c, middle panel). Mitochondrial decrease in ROS levels in transformed cells (Figure 6h). The morphological analysis, performed in both cell lines grown in addition to the cultivation medium of N-acetyl-cysteine (NAC), 25 mM glucose, resulted in comparable data (Supplementary which decreased the ROS levels in both normal and transformed Figure S5). A role of PKA on mitochondrial dynamics of cells, partly sustained the growth of transformed cells but had no transformed cells was confirmed by analysis of mitochondrial effect on normal cells (Supplementary Figure S6). FSK-treated morphology upon co-treatment of the cells with FSK and H89. As reverted cells did not show ATP levels increase (Supplementary

& 2013 Macmillan Publishers Limited Oncogene (2013) 352 --362 PKA regulation of K-ras cancer cell mitochondria R Palorini et al 356

Figure 5. FSK treatment increases mitochondrial interconnections in transformed cells. (a) Confocal microscopy images of stable EYFP-Mito expressing clones of normal and transformed cells sorted according their mitochondrial morphology: fragmented, intermediate and networked. (b, c) Mitochondrial morphology analysis was carried out at indicated time points in normal (b) and transformed (c) cells grown in 1mM glucose, subjected or not to treatment with FSK and 2 mM H89. For each determination, at least 100 cells were counted and classiﬁed depending their mitochondrial morphology. Data represent the average of at least three independent experiments (±s.d.) and are indicated as percentage.

Figure S2D) and ROS levels decrease (Supplementary Figure S2E). harboring an oncogenic K-ras, namely MIA PaCa-2 (pancreatic) Noteworthy, the positive effect on complex I by FSK treatment in and A549 (lung). In fact, as shown in Supplementary Figure S8, transformed cells was observed also when they were grown in both were sensible to glucose deprivation, undergoing to cell 25 mM glucose (Supplementary Figure S7), pointing out to a death, and both were, in different extend, protected by FSK general role of PKA as mitochondrial function regulator. treatment. After 24 h of FSK treatment, MDA-MB-231 cells showed an increase in cellular cAMP levels (Figure 8a), in complex I activity (Figure 8b) and in ATP levels (Figure 8c and Supplementary Figure Mitochondrial dysfunctions in breast cancer MDA-MB-231 cells S9A) associated with a decrease in ROS levels (Figure 8d and and FSK treatment effects Supplementary Figure S9B) and an increase in mitochondrial MDA-MB-231 cells, harboring a K-ras oncogenic mutation, show a interconnections (Figure 8f). The increase in mitochondrial cellular metabolism similar to those observed in the mouse model interconnections (Figure 8f, right panel) and ATP levels (data not 54 of K-ras-dependent transformation. Mitochondrial and metabolic shown) produced by FSK were prevented by H89 treatment, characteristics of these human cancer cells were, thus, analyzed, conﬁrming the role of PKA in mitochondria regulation. Note- both in untreated and FSK-treated conditions. MDA-MB-231 cells, worthy, the FSK-dependent positive effects on complex I activity grown at 1 mM glucose, 25 mM galactose (Figure 7a) as well as at and mitochondrial fusion were observed also in MDA-MB-231 cells 5mM fructose (Supplementary Figure S1C) showed a dramatic grown in 25 mM glucose (Supplementary Figure S10) as well as decrease in their proliferation capacity as compared with the cells previously described for mouse transformed cells. grown at 25 mM glucose (Figure 7a). Addition of OM in both glucose conditions did not cause any effect on total ATP levels (Figure 7b), showing their dependence on glycolytic activity. FSK Mitochondrial fusion is induced also by short-term FSK treatment treatment of MDA-MB-231 cells grown at 1 mM glucose rescued PKA-catalyzed phosphorylation of cellular proteins can result, in their proliferation ability and decreased apoptosis (Figures 7c--e). addition to long-term transcriptional effects as those mediated by Similar results were obtained in other two human cancer cell lines CREB, in short-term regulation of mitochondrial respiration and

Oncogene (2013) 352 --362 & 2013 Macmillan Publishers Limited PKA regulation of K-ras cancer cell mitochondria R Palorini et al 357 ab DISCUSSION Among the various cellular systems perturbed by oncogenic K-ras mutations in different cellular models, derangement in the cAMP/PKA signal pathway and its target systems has been reported.24,31,64 The oncogenic K-ras-transformed cells analyzed in this study exhibit a strict dependence on high glucose concentration for growth. This shift from aerobic to fermentative ATP supply is cd shown, here, to be associated with a derangement in the cAMP/ PKA system, resulting in structural and functional alterations of mitochondria. K-ras-transformed mouse fibroblasts show a higher content of cAMP as compared with control fibroblasts. This and the finding that in these transformed cells, as well as in K-ras human breast cancer MDA-MB-231 cell line, adenyl cyclase activation by FSK results in a much larger increase in the cAMP level as compared with normal cells confirm an increased cyclase ef responsiveness to stimuli. The enhanced tendency of K-ras- transformed fibroblasts to produce cAMP was, however, found to be associated to a reduced enzymatic activity of PKA and reduced phosphorylation of its substrate CREB, as compared with normal fibroblasts. Both these deficits were rescued by elevation of cAMP levels induced by FSK, which also prevented death of some human transformed K-ras cells like MDA-MB-231, MIA PaCa-2 and A549. Direct immunochemical determination of the catalytic gh subunit of PKA in transformed and normal mouse fibroblasts showed that the differences in the functional capacity of PKA between the two cell lines were not due to changes in the content of the enzyme but reflected altered responsiveness of PKA to cellular activators/inhibitors. The deregulation of the cAMP/PKA pathway can possibly represent a feed-back compensatory response of the cyclase to the depression of PKA activity. Figure 6. FSK treatment improves mitochondrial activity especially The large increase in the cellular cAMP level induced by FSK in transformed cells. The analyses were performed at 72 h of culture and the resulting activation of PKA activity, with promotion of at 1 mM glucose (Glc) in absence or in presence for 48 h of FSK or phosphorylation of PKA substrates like CREB and Drp1, prevented FSK plus 2 mM H89 as indicated. (a, b) Complex I activity of death of mouse and human ras-transformed cells in 1 mM glucose- mitochondrial respiratory chain was measured in daily isolated limiting conditions. mitoplast fraction from normal (a) and transformed (b) cells; K-ras-transformed fibroblasts show depression of complex I # Po0.02, Student’s t-test. (c, d) Complex IV activity of mitochondrial activity and increase in ROS levels as compared with normal respiratory chain was measured in daily isolated mitoplast fraction fibroblasts. As previously observed in various cellular pathophy- from normal (c) and transformed (d) cells. Total intracellular ATP was siological conditions37,50 elevation of the cAMP level, induced by measured in normal (e) and transformed (f) cells; *Po0.0001 and # FSK in mouse and human ras-transformed cells, promoted the Po0.02, Student’s t-test. (g, h) ROS levels were measured in normal (g) and transformed (h) cells; **Po0.003, Student’s t-test. All data activity of complex I and decreased the ROS levels. Depression of represent the average of at least three independent experiments complex I, which is by itself a major producer of ROS and is (±s.d.). particularly vulnerable to oxidative stress,65 can set up a ‘vicious’ cycle, detrimental for cell growth. In fact, decrease in ROS levels, affected by the addition of N-acetyl-cysteine, did promote the morphology at post-translational level.39,50,63 As reported for growth of ras-transformed cells. K-ras-transformed fibroblasts, no significant time-dependent CREB It is worth noting that the activity of complex IV was not phosphorylation was observed in MDA-MB-231 cells grown at depressed in K-ras-transformed fibroblasts, neither was it affected either 25 or 1 mM glucose (Figures 9a and b). The addition of FSK, by FSK addition. The activity of complex IV has been reported to however, resulted in a transient increase in CREB phosphorylation, be affected by phosphorylation of its subunits.39 In the cell lines which after reaching a maximum at 1 h declined back to the initial and the prevailing experimental conditions of the present work, value at 24 h of cultivation (Figures 9c and d). the functional activity of complex IV is stable and unaffected by It has been reported that PKA-mediated phosphorylation of eventual phosphorylation (see also Piccoli et al.).50 the pro-fission protein factor Drp1 inhibits fission with consequent The cAMP/PKA system is involved in regulation of the dynamics increase in mitochondrial fusion. The results presented in of mitochondrial fusion/fission.40,41,66 In K-ras-transformed cells Figure 10a show that FSK treatment of MDA-MB-231 cells induced mitochondria were largely in a fragmented state. Elevation of the phosphorylation of Drp1 between 1 and 4 h, which then cAMP level by FSK largely reverted mitochondrial fragmentation, disappeared at 24-h cultivation. an effect that was prevented by the PKA inhibitor H89. The effects of FSK and of the mitochondrial division inhibitor Importantly, the positive effects of FSK treatment on the Mdivi-1, which prevents mitochondrial fission by inhibiting the mitochondrial morphology and activity were observed both in Drp1 activity, were analyzed on mitochondrial dynamics in short- high and low glucose availability implying that PKA pathway time intervals. As shown in Figure 10b, treatment with FSK or regulates mitochondrial function, at least in our experimental Mdivi-1 of MDA-MB-231 cells grown at 1 mM glucose significantly conditions, independently from glucose deprivation. It appears as increased mitochondrial interconnections as early as 1-h treat- the responsiveness to cAMP of the down-stream components of ment. Analysis of ROS levels showed that the more interconnected signal transduction is reduced in K-ras mutated cells, just opposite mitochondria induced by FSK or Mdivi-1 were associated with a to the increased excitability of the cyclase. Altogether, these significant reduction of ROS levels (Figure 10c). results show that the derangement of cAMP/PKA pathway in

& 2013 Macmillan Publishers Limited Oncogene (2013) 352 --362 PKA regulation of K-ras cancer cell mitochondria R Palorini et al 358 a b

cde

Figure 7. FSK treatment protects human MDA-MB-231 cancer cells from glucose shortage-dependent cell death. (a) Proliferation curves of MDA-MB-231 cells cultured at 25 mM glucose (Glc), at 1 mM Glc or at 25 mM galactose (Gal) were determined counting cells at indicated time points. (b) Total intracellular ATP was measured in MDA-MB-231 cells grown at 25 and 1 mM Glc, treated or not with 5 mM OM at 24 h of culture. Values are relative to the sample 25 mM Glc. (c) Proliferation curves of MDA-MB-231 cells grown at 1 mM Glc, treated or not with FSK, were determined counting cells at indicated time points. (d) Phase contrast microscopy images were collected for cells at 72 h of culture. (e) Analysis of cell viability in MDA-MB-231 cells was evaluated by annexin-V/propidium iodide (PI) staining and calculating the percentage of annexin-V/PI-negative cells. All data represent the average of at least three independent experiments (±s.d.).

ras-transformed cells results in mitochondrial structure and and 100 mg/ml streptomycin (complete medium), supplemented with 10% respiration alterations with consequent cells dependence on newborn calf serum (mouse cells) or 5--10% fetal bovine serum (human glucose availability. cells). All reagents for media were purchased from Invitrogen (Carlsbad, The relationship between the dynamic equilibrium between CA, USA). fusion and fission of mitochondria and their respiratory activity is a 67 matter of current investigation. A set of G proteins has been Cell treatments found to regulate mitochondrial dynamics.68 Among these the For long-term studies, cells were plated at the density of 3000, 5000 (MIA Drp1 protein promotes mitochondrial fission. Inactivation of Drp1, 2 effected by its phosphorylation by PKA, can result in promotion of PaCa-2) or 15 000 (A549) cells/cm in complete growth medium. After 16 h, fusion and hence in protection from autophagy.63 Immunochem- cells were washed twice with phosphate-buffered saline and incubated in ical assays showed that in MDA-MB-231 cells, whose mitochondria growth medium (time 0) without glucose and sodium pyruvate (Invitrogen), were prevalently in the fragmented state, Drp1 was essentially in supplemented with 25 or 1 mM glucose, 25 mM galactose or 5 mM fructose. dephosphorylated state. Addition of FSK, which also in this case Glucose, galactose and fructose were purchased from Sigma-Aldrich Inc. reverted mitochondrial fragmentation, promoted phosphorylation (St Louis, MO, USA). To stimulate the increase in cAMP levels, cells were of Drp1. Other experiments showed that suppression of the daily treated with 10 mM forskolin (FSK) (Sigma-Aldrich) starting from 24 h activity of Drp-1 by its antagonist Mdivi-1 increased mitochondrial after medium replacement. Cells were then collected for further analyses at interconnection in MDA-MB-231 cells. Recent observations sup- 24, 48, 72, 96 and in some cases at 120 h of culture. Wherein indicated, to port the link between respiratory chain complexes activity, in specifically inhibit PKA, H89 (N-[2-p-bromocinnamylamino)ethyl]-5-isoqui- particular complex I, and the mitochondrial fusion/fission pro- nolinesulfonamide dihydrochloride from Sigma-Aldrich) was used in cess.69,70 Conversely, inhibition of complex I by rotenone results in combination with FSK. For short-term studies, cells were cultured as mitochondrial fission.71 above described, except that they were treated at a single time point and In conclusion, this study shows a derangement of the cAMP/ then analyzed. PKA pathway in oncogenic K-ras-transformed cells, which results in impairment of mitochondrial fusion, associated with decreased Intracellular ATP quantification activity of complex I, increased ROS production and apoptosis. Intracellular ATP levels were measured using CellTiter Glo luciferin- These observations might serve to design new approaches in the luciferase assay (Promega, Madison, WI, USA) as described in Gaglio et al.54 development of anticancer drugs. Wherein indicated, intracellular ATP levels were measured after cells treatment with 5 mM OM (Sigma-Aldrich).

MATERIALS AND METHODS Western blot analysis Cell cultures For the analysis of total CREB, phospho-CREB Ser133, catalytic subunit of Mouse ﬁbroblast NIH3T3 cells (ATCC, Manassas, VA, USA), K-ras- PKA, Drp1 and phospho-Drp1 Ser637 levels, 2 Â 105 cells were harvested 72 transformed NIH3T3-derived cell line 226.4.1, reverted NIH3T3, MDA- and lysed in sodium dodecyl sulfate sample buffer 1 Â . Samples were then MB-231, MIA PaCa-2 and A549 cells were routinely cultured in Dulbecco’s resolved by sodium dodecyl sulfate--polyacrylamide gel electrophoresis modiﬁed Eagle’s medium containing 4 mML-glutamine, 100 U/ml penicillin and transferred to nitrocellulose membrane, which was incubated

Oncogene (2013) 352 --362 & 2013 Macmillan Publishers Limited PKA regulation of K-ras cancer cell mitochondria R Palorini et al 359

Figure 8. FSK treatment enhances several mitochondrial parameters in MDA-MB-231 cancer cells associated to increased survival on glucose shortage. cAMP levels (a), complex I activity (b), ATP (c) and ROS (d) levels were measured at 48 h of culture in MDA-MB-231 cells grown at 1mM glucose (Glc). Panel (a) represents the mean ±s.e.m. of three independent determinations; all the other data represent the average of at least three independent experiments (±s.d.); #Po0.02, *Po0.0001 and **Po0.003, Student’s t-test. (e) Examples of fragmented, intermediate and networked mitochondria in stable EYFP-Mito expressing MDA-MB-231 cells. (f) Analysis of mitochondrial morphology in MDA-MB-231 cells grown at 1 mM Glc, subjected or not to treatment with FSK and 5 mM H89. For each determination, at least 100 cells were counted and classiﬁed depending their mitochondrial morphology. Data represent the average of at least three independent experiments (±s.d.) and are indicated as percentage; *Po0.0001 and #Po0.05, Student’s t-test; P-value is calculated on percentage of fragmented and networked mitochondria. FSK-treated samples are compared with untreated samples, while samples treated with FSK þ H89 are compared with FSK- treated samples.

a b

Figure 9. Expression analysis of total CREB and phospho-CREB Ser133 in MDA-MB-231 cells. (a) MDA-MB-231 cells grown at 25 or 1 mM glucose (Glc) were collected at indicated time points and total cellular extracts were subjected to sodium dodecyl sulfate--polyacrylamide gel electrophoresis followed by western blot analysis with speciﬁc antibodies against total CREB (Creb) and phospho-CREB Ser133 (P-Creb). (b) Quantitative analysis of CREB phosphorylation status was performed by densitometric analysis of western blot ﬁlms. The values obtained for P-CREB were normalized to the corresponding total CREB values, plotted as fold change from the sample 0 h (0 h ¼ 1) and indicated as relative densitometric units (DU). (c, d) Western blot evaluation (c) of phospho-CREB Ser133 was also performed in MDA-MB-231 cells grown at 1 mM glucose on 24 h of FSK treatment and quantitative analysis is reported in the histogram (d). Values were normalized as described above. All data represent the mean ±s.e.m. of three independent determinations.

& 2013 Macmillan Publishers Limited Oncogene (2013) 352 --362 PKA regulation of K-ras cancer cell mitochondria R Palorini et al 360 a cytometric data was carried out using the freely available WinMDI software. N-acetyl-L-cysteine was purchased from Sigma-Aldrich.

Mitochondrial morphology analysis To study mitochondrial morphology, NIH3T3, K-ras-transformed NIH3T3 bcand MDA-MB-231 cells were transfected with pEYFP-Mito construct (6115-1, BD Biosciences) and stably transfected clones were isolated. Images of mitochondrial morphology were collected under a Nikon ECLIPSE 90i fluorescence microscope (Nikon, Tokyo, Japan) equipped with a b/w CCD camera (Hamamatsu-CoolSnap, Hamamatsu Corporation, Hamamatsu City, Japan) or using a laser scanning confocal microscope MRC-600 (Bio-Rad Microscience Division, Hemel Hempstead, UK) coupled to an Optiphot-2 Epi Fluor microscope (Nikon). Images were then visualized, processed and classified by using the freely available ImageJ software. When indicated, to selectively induce mitochondrial interconnections Figure 10. Networked mitochondria are less prone to ROS genera- 10 mM Mdivi-1 (3-(2,4-dichloro-5-methoxyphenyl)-2,3-dihydro-2--thioxo- tion. (a) Analysis of Total Drp1 and phospho-Drp1 Ser637 (p-drp1) 4(1H)-quinazolinone from Sigma-Aldrich) was used. was performed in MDA-MB-231 cells grown at 1 mM glucose untreated (À) or treated with FSK for 1, 4 and 24 h. Cells were collected at indicated time points and total cellular extracts were D-glucose and L-lactate measurement subjected to sodium dodecyl sulfate--polyacrylamide gel electro- D-glucose and L-lactate levels in culture medium were determined as phoresis followed by western blot analysis with specific antibodies described in Chiaradonna et al.52 against Drp1 and phospho-Drp1. (b, c) MDA-MB-231 cells, cultured at 1 mM glucose, were subjected to a treatment with 10 mM FSK or 10 mM Mdivi-1 48 h after medium change. Cells were then collected Glucose uptake 1 h and 4 h after the treatment for the subsequent analyses. (b) Mitochondrial morphology was determined classifying mitochon- Glucose uptake assay was performed using 2-NBDG (Invitrogen) fluor- dria as fragmented, intermediate and networked (for major details escent glucose analog. The cells were washed twice with phosphate- of the procedure refer to Figure 8 legend). (c) Intracellular ROS levels buffered saline, then their medium was replaced with a glucose-free were analyzed staining cells with DCFH2-DA. Data represent the medium supplemented with 60 mM 2-NBDG for 30 min at 37 1C. After average of three independent experiments (±s.d.). staining, cells were trypsinized, collected in phosphate-buffered saline with 10% newborn calf serum, acquired with FACScan flow cytometer and analyzed with WinMDI software. overnight with specific antibodies (total CREB, phospho-CREB Ser133 and phospho-Drp1 Ser637 were obtained from Cell Signaling Technology Inc., Danvers, MA, USA; total Drp1 from BD Biosciences, Franklin Lakes, NJ, USA; CONFLICT OF INTEREST catalytic subunit of PKA and b-actin from Santa Cruz Biotechnology Inc., The authors declare no conflict of interest. Santa Cruz, CA, USA).

ACKNOWLEDGEMENTS PKA activity This work has been supported by a grant to LA from MIUR (FIRB-ITALBIONET), grants Vmax was obtained from Lineweaver-Burk plots of enzymatic PKA activity. to FC from the Italian Government (FAR) and MIUR (Prin 2008), grant to DDR and CC In all, 15 mg proteins of sonicated cellular extract were incubated in 50 mlof from MIUR (Progetto FIRB futuro in ricerca, 2008), grants to SP from MIUR (Progetto 10 mM Tris--HCl, pH 7.5, 8 mM MgCl2,20mM NaF, 0.25 mM phenylmethyl- FIRB Rete Nazionale per lo Studio della Proteomica Umana-Italian Human sulfonyl fluoride and 3 mg of OM in the presence of [g-32P]ATP (1000 c.p.m./ ProteomeNet, 2009) and Progetto Strategico Ric.002, Cip PS 101, POR 2000/06. pmol), 1 mM cAMP and histone H2B (2.5 mg), as substrate, at different ATP concentration (30--190 mM). After 15 min at 30 1C, proteins were subjected to sodium dodecyl sulfate--polyacrylamide gel electrophoresis. After REFERENCES electrophoresis, the gels were stained with Comassie blue and dried. 1 Mathupala SP, Rempel A, Pedersen PL. Aberrant glycolytic metabolism of cancer Radioactive protein bands of phosphorylated histone were detected by cells: a remarkable coordination of genetic, transcriptional, post-translational, and Personal FX at ‘phosphorus imager’ (Bio-Rad, Hercules, CA, USA) and mutational events that lead to a critical role for type II hexokinase. J Bioenerg quantified by VERSADOC (Bio-Rad). Biomembr 1997; 29: 339 --343. 2 Mazurek S, Eigenbrodt E. The tumor metabolome. Anticancer Res 2003; 23: 1149 --1154. cAMP cellular levels measurement and OXPHOS complex I and 3 Zu XL, Guppy M. Cancer metabolism: facts, fantasy, and fiction. Biochem Biophy complex IV activities determination Res Commun 2004; 313: 459 --465. These assays were performed with little modifications as described in 4 Ramanathan A, Wang C, Schreiber SL. Perturbational profiling of a cell-line model De Rasmo et al.38 of tumorigenesis by using metabolic measurements. Proc Natl Acad Sci USA 2005; 102: 5992 --5997. 5 Moreno-Sanchez R, Rodriguez-Enriquez S, Marin-Hernandez A, Saavedra E. Energy Flow cytometric analyses metabolism in tumor cells. FEBS J 2007; 274: 1393 --1418. Detection of cell viability and ROS levels was carried out through flow 6 McFate T, Mohyeldin A, Lu H, Thakar J, Henriques J, Halim ND et al. Pyruvate dehydrogenase complex activity controls metabolic and malignant phenotype in cytometric analyses using a FACScan flow cytometer (Becton-Dickinson, cancer cells. J Biol Chem 2008; 283: 22700 --22708. Franklin Lakes, NJ, USA) with CellQuest software (Becton-Dickinson). Cell 7 Warburg O. On the origin of cancer cells. Science 1956; 123: viability was evaluated using the apoptosis assay kit (EXBIO Antibodies, 309 --314. Prague, Czech Republic) as specified in the manufacturer’s datasheet. ROS 8 Vander Heiden MG, Cantley LC, Thompson CB. Understanding the Warburg effect: levels were determined as described in Gaglio et al.54 Analysis of flow the metabolic requirements of cell proliferation. Science 2009; 324: 1029 --1033.

Oncogene (2013) 352 --362 & 2013 Macmillan Publishers Limited PKA regulation of K-ras cancer cell mitochondria R Palorini et al 361 9 Chiaradonna F, Moresco RM, Airoldi C, Gaglio D, Palorini R, Nicotra F et al. From 36 Papa S. Does cAMP play a part in the regulation of the mitochondrial electron cancer metabolism to new biomarkers and drug targets. Biotechnol Adv 2011; 30: transport chain in mammalian cells? IUBMB Life 2006; 58: 173 --175. 30 --51. 37 Papa S, De Rasmo D, Scacco S, Signorile A, Technikova-Dobrova Z, Palmisano G 10 Tortora G, Ciardiello F. Protein kinase A as target for novel integrated strategies of et al. Mammalian complex I: a regulable and vulnerable pacemaker in cancer therapy. Ann NY Acad Sci 2002; 968: 139 --147. mitochondrial respiratory function. Biochimica et Biophysica Acta 2008; 1777: 11 Shaw RJ, Cantley LC. Ras, PI(3)K and mTOR signalling controls tumour cell growth. 719 --728. Nature 2006; 441: 424 --430. 38 De Rasmo D, Gattoni G, Papa F, Santeramo A, Pacelli C, Cocco T et al. The 12 Ma W, Sung HJ, Park JY, Matoba S, Hwang PM. A pivotal role for p53: balancing beta-adrenoceptor agonist isoproterenol promotes the activity of respiratory aerobic respiration and glycolysis. J Bioenerg Biomembr 2007; 39: 243 --246. chain complex I and lowers cellular reactive oxygen species in fibroblasts and 13 Chiaradonna F, Gaglio D, Vanoni M, Alberghina L. Expression of transforming heart myoblasts. Eur J Pharmacol 2011; 652:15--22. K-Ras oncogene affects mitochondrial function and morphology in mouse 39 Acin-Perez R, Salazar E, Kamenetsky M, Buck J, Levin LR, Manfredi G. Cyclic AMP fibroblasts. Biochimica et Biophysica Acta 2006; 1757: 1338 --1356. produced inside mitochondria regulates oxidative phosphorylation. Cell Metabol 14 Lu J, Sharma LK, Bai Y. Implications of mitochondrial DNA mutations and 2009; 9: 265 --276. mitochondrial dysfunction in tumorigenesis. Cell Res 2009; 19: 802 --815. 40 Cribbs JT, Strack S. Reversible phosphorylation of Drp1 by cyclic AMP-dependent 15 Lee HC, Wei YH. Mitochondrial DNA instability and metabolic shift in human protein kinase and calcineurin regulates mitochondrial fission and cell death. cancers. Int J Mol Sci 2009; 10: 674 --701. EMBO Rep 2007; 8: 939 --944. 16 Kim JW, Tchernyshyov I, Semenza GL, Dang CV. HIF-1-mediated expression of 41 Chang CR, Blackstone C. Cyclic AMP-dependent protein kinase phosphorylation of pyruvate dehydrogenase kinase: a metabolic switch required for cellular Drp1 regulates its GTPase activity and mitochondrial morphology. J Biol Chem adaptation to hypoxia. Cell Metabol 2006; 3: 177 --185. 2007; 282: 21583 --21587. 17 Buchet K, Godinot C. Functional F1-ATPase essential in maintaining growth and 42 Bellis A, Castaldo D, Trimarco V, Monti MG, Chivasso P, Sadoshima J et al. membrane potential of human mitochondrial DNA-depleted rho degrees cells. Cross-talk between PKA and Akt protects endothelial cells from apoptosis J Biol Chem 1998; 273: 22983 --22989. in the late ischemic preconditioning. Arterioscler Thromb Vasc Biol 2009; 29: 18 Simonnet H, Demont J, Pfeiffer K, Guenaneche L, Bouvier R, Brandt U et al. 1207 --1212. Mitochondrial complex I is deficient in renal oncocytomas. Carcinogenesis 2003; 43 Harada H, Becknell B, Wilm M, Mann M, Huang LJ, Taylor SS et al. Phosphorylation 24: 1461 --1466. and inactivation of BAD by mitochondria-anchored protein kinase A. Mol Cell 19 Alchanati I, Nallar SC, Sun P, Gao L, Hu J, Stein A et al. A proteomic analysis reveals 1999; 3: 413 --422. the loss of expression of the cell death regulatory gene GRIM-19 in human renal 44 Zippin JH, Chen Y, Nahirney P, Kamenetsky M, Wuttke MS, Fischman DA et al. cell carcinomas. Oncogene 2006; 25: 7138 --7147. Compartmentalization of bicarbonate-sensitive adenylyl cyclase in distinct 20 Plecita-Hlavata L, Lessard M, Santorova J, Bewersdorf J, Jezek P. Mitochondrial signaling microdomains. FASEB J 2003; 17:82--84. oxidative phosphorylation and energetic status are reflected by morphology of 45 Chen Q, Lin RY, Rubin CS. Organelle-specific targeting of protein kinase AII (PKAII). mitochondrial network in INS-1E and HEP-G2 cells viewed by 4Pi microscopy. Molecular and in situ characterization of murine A kinase anchor proteins that Biochimica et Biophysica Acta 2008; 1777: 834 --846. recruit regulatory subunits of PKAII to the cytoplasmic surface of mitochondria. 21 Bos JL. Ras oncogenes in human cancer: a review. Cancer Res 1989; 49: J Biol Chem 1997; 272: 15247 --15257. 4682--4689. 46 Schwoch G, Trinczek B, Bode C. Localization of catalytic and regulatory subunits of 22 Rodenhuis S. Ras and human tumors. Semin Cancer Biol 1992; 3: 241 --247. cyclic AMP-dependent protein kinases in mitochondria from various rat tissues. 23 Franks DJ, Whitfield JF, Durkin JP. Viral p21 Ki-RAS protein: a potent intracellular Biochem J 1990; 270: 181 --188. mitogen that stimulates adenylate cyclase activity in early G1 phase of cultured 47 Signorile A, Sardanelli AM, Nuzzi R, Papa S. Serine (threonine) phosphatase(s) rat cells. J Cell Biochem 1987; 33:87--94. acting on cAMP-dependent phosphoproteins in mammalian mitochondria. FEBS 24 Konishi-Imamura L, Tohda M, Noda M, Nakano H, Nomura Y. GTP-binding proteins Lett 2002; 512: 91 --94. and adenylate cyclase activity in v-Ki-ras transformed NIH/3T3 fibroblast cells. 48 Sardanelli AM, Signorile A, Nuzzi R, Rasmo DD, Technikova-Dobrova Z, Drahota Z Biochem Biophy Res Commun 1988; 153: 1214 --1222. et al. Occurrence of A-kinase anchor protein and associated cAMP-dependent 25 Konishi-Imamura L, Noda M, Nomura Y. Alteration by v-Ki-ras in NaF, cholera toxin protein kinase in the inner compartment of mammalian mitochondria. FEBS Lett and forskolin-induced adenylate cyclase activation in NIH/3T3 fibroblast cells. 2006; 580: 5690 --5696. Biochem Biophy Res Commun 1987; 146:47--52. 49 De Rasmo D, Signorile A, Roca E, Papa S. cAMP response element-binding protein 26 Gallo A, Feliciello A, Varrone A, Cerillo R, Gottesman ME, Avvedimento VE. Ki-ras (CREB) is imported into mitochondria and promotes protein synthesis. FEBS J oncogene interferes with the expression of cyclic AMP-dependent promoters. Cell 2009; 276: 4325 --4333. Growth Differ 1995; 6: 91 --95. 50 Piccoli C, Scacco S, Bellomo F, Signorile A, Iuso A, Boffoli D et al. cAMP controls 27 Feliciello A, Giuliano P, Porcellini A, Garbi C, Obici S, Mele E et al. The v-Ki-Ras oxygen metabolism in mammalian cells. FEBS Lett 2006; 580: 4539 --4543. oncogene alters cAMP nuclear signaling by regulating the location and the 51 Bossu P, Vanoni M, Wanke V, Cesaroni MP, Tropea F, Melillo G et al. A dominant expression of cAMP-dependent protein kinase IIbeta. J Biol Chem 1996; 271: negative RAS-specific guanine nucleotide exchange factor reverses neoplastic 25350 --25359. phenotype in K-ras transformed mouse fibroblasts. Oncogene 2000; 19: 28 Ross DT, Scherf U, Eisen MB, Perou CM, Rees C, Spellman P et al. Systematic 2147 --2154. variation in gene expression patterns in human cancer cell lines. Nat Genet 2000; 52 Chiaradonna F, Sacco E, Manzoni R, Giorgio M, Vanoni M, Alberghina L. Ras- 24: 227 --235. dependent carbon metabolism and transformation in mouse fibroblasts. 29 Chiaradonna F, Balestrieri C, Gaglio D, Vanoni M. RAS and PKA pathways in cancer: Oncogene 2006; 25: 5391 --5404. new insight from transcriptional analysis. Front Biosci 2008; 13: 5257 --5278. 53 Gaglio D, Soldati C, Vanoni M, Alberghina L, Chiaradonna F. Glutamine deprivation 30 Baratta MG, Porreca I, Di Lauro R. Oncogenic ras blocks the cAMP pathway and induces abortive s-phase rescued by deoxyribonucleotides in k-ras transformed dedifferentiates thyroid cells via an impairment of pax8 transcriptional activity. fibroblasts. PLoS One 2009; 4: e4715. Mol Endocrinol 2009; 23: 838 --848. 54 Gaglio D, Metallo CM, Gameiro PA, Hiller K, Sala Danna L, Balestrieri C et al. 31 Balestrieri C, Alberghina L, Vanoni M, Chiaradonna F. Data recovery and Oncogenic K-Ras decouples glucose and glutamine metabolism to support cancer integration from public databases uncovers transformation-specific transcrip- cell growth. Mol Syst Biol 2011; 7:523. tional downregulation of cAMP-PKA pathway-encoding genes. BMC Bioinform 55 Baracca A, Chiaradonna F, Sgarbi G, Solaini G, Alberghina L, Lenaz G. 2009; 10(Suppl 12): S1. Mitochondrial complex I decrease is responsible for bioenergetic dysfunction in 32 Hiraki Y, McMorrow IM, Birnbaum MJ. The regulation of glucose transporter gene K-ras transformed cells. Biochimica et Biophysica Acta 2010; 1797: 314 --323. expression by cyclic adenosine monophosphate in NIH3T3 fibroblasts. Mol 56 Yamamoto F, Perucho M. Activation of a human c-K-ras oncogene. Nucleic Acids Endocrinol 1989; 3: 1470 --1476. Res 1984; 12: 8873 --8885. 33 Hosaka Y, Tawata M, Kurihara A, Ohtaka M, Endo T, Onaya T. The regulation of two 57 Yun J, Rago C, Cheong I, Pagliarini R, Angenendt P, Rajagopalan H et al. Glucose distinct glucose transporter (GLUT1 and GLUT4) gene expressions in cultured rat deprivation contributes to the development of KRAS pathway mutations in tumor thyroid cells by thyrotropin. Endocrinology 1992; 131: 159 --165. cells. Science 2009; 325: 1555 --1559. 34 Osawa H, Printz RL, Whitesell RR, Granner DK. Regulation of hexokinase II gene 58 Vizan P, Boros LG, Figueras A, Capella G, Mangues R, Bassilian S et al. K-ras transcription and glucose phosphorylation by catecholamines, cyclic AMP, and codon-specific mutations produce distinctive metabolic phenotypes in NIH3T3 insulin. Diabetes 1995; 44: 1426 --1432. mice [corrected] fibroblasts. Cancer Res 2005; 65: 5512 --5515. 35 Depre C, Ponchaut S, Deprez J, Maisin L, Hue L. Cyclic AMP suppresses the 59 Weinberg F, Hamanaka R, Wheaton WW, Weinberg S, Joseph J, Lopez M et al. inhibition of glycolysis by alternative oxidizable substrates in the heart. J Clin Mitochondrial metabolism and ROS generation are essential for Kras-mediated Invest 1998; 101: 390 --397. tumorigenicity. Proc Natl Acad Sci USA 2010; 107: 8788 --8793.

& 2013 Macmillan Publishers Limited Oncogene (2013) 352 --362 PKA regulation of K-ras cancer cell mitochondria R Palorini et al 362 60 Gerits N, Kostenko S, Shiryaev A, Johannessen M, Moens U. Relations between the 66 Merrill RA, Dagda RK, Dickey AS, Cribbs JT, Green SH, Usachev YM et al. mitogen-activated protein kinase and the cAMP-dependent protein kinase Mechanism of neuroprotective mitochondrial remodeling by PKA/AKAP1. PLoS pathways: comradeship and hostility. Cell Signal 2008; 20: 1592 --1607. Biol 2011; 9: e1000612. 61 Papa S, Zanotti F, Cocco T, Perrucci C, Candita C, Minuto M. Identification 67 Grandemange S, Herzig S, Martinou JC. Mitochondrial dynamics and cancer. of functional domains and critical residues in the adenosinetriphosphatase Semin Cancer Biol 2009; 19:50--56. inhibitor protein of mitochondrial F0F1 ATP synthase. Eur J Biochem 1996; 240: 68 Palmer CS, Osellame LD, Stojanovski D, Ryan MT. The regulation of mitochondrial 461 --467. morphology: intricate mechanisms and dynamic machinery. Cell Signal 2011; 23: 62 Chijiwa T, Mishima A, Hagiwara M, Sano M, Hayashi K, Inoue T et al. Inhibition of 1534 --1545. forskolin-induced neurite outgrowth and protein phosphorylation by a newly 69 Koopman WJ, Visch HJ, Verkaart S, van den Heuvel LW, Smeitink JA, Willems PH. synthesized selective inhibitor of cyclic AMP-dependent protein kinase, Mitochondrial network complexity and pathological decrease in complex I activity N-[2-(p-bromocinnamylamino)ethyl]-5-isoquinolinesulfonamide (H-89), of PC12D are tightly correlated in isolated human complex I deficiency. Am J Physiol Cell pheochromocytoma cells. J Biol Chem 1990; 265: 5267 --5272. Physiol 2005; 289: C881 --C890. 63 Gomes LC, Di Benedetto G, Scorrano L. During autophagy mitochondria 70 Capaldi RA, Aggeler R, Gilkerson R, Hanson G, Knowles M, Marcus A et al. A elongate, are spared from degradation and sustain cell viability. Nat Cell Biol replicating module as the unit of mitochondrial structure and functioning. 2011; 13: 589 --598. Biochimica et Biophysica Acta 2002; 1555: 192 --195. 64 Borlikova G, Endo S. Inducible cAMP early repressor (ICER) and brain functions. 71 Benard G, Bellance N, James D, Parrone P, Fernandez H, Letellier T et al. Mol Neurobiol 2009; 40:73--86. Mitochondrial bioenergetics and structural network organization. J Cell Sci 2007; 65 Lenaz G, Baracca A, Barbero G, Bergamini C, Dalmonte ME, Del Sole M et al. 120(Pt 5): 838 --848. Mitochondrial respiratory chain super-complex I-III in physiology and pathology. 72 Pulciani S, Santos E, Long LK, Sorrentino V, Barbacid M. ras gene Amplification and Biochimica et Biophysica Acta 2010; 1797: 633 --640. malignant transformation. Mol Cell Biol 1985; 5: 2836 --2841.

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

Oncogene (2013) 352 --362 & 2013 Macmillan Publishers Limited