Mitochondrial localization of TIGAR under hypoxia stimulates HK2 and ...

Download PDF
1 downloads 21 Views 1MB Size Report
Comment
Dec 11, 2012 - Edited by Tak W. Mak, The Campbell Family Institute for Breast Cancer Research, Ontario Cancer Institute at Princess Margaret Hospital, ...
Mitochondrial localization of TIGAR under hypoxia stimulates HK2 and lowers ROS and cell death Eric C. Cheung, Robert L. Ludwig, and Karen H. Vousden1 The Beatson Institute for Cancer Research, Switchback Road, Glasgow G61 1BD, United Kingdom Edited by Tak W. Mak, The Campbell Family Institute for Breast Cancer Research, Ontario Cancer Institute at Princess Margaret Hospital, University Health Network, Toronto, ON, Canada, and approved November 2, 2012 (received for review April 19, 2012)

cancer metabolism

| glycolysis | oxidative stress | low oxygen

T

he uptake and metabolism of glucose is fundamental for energy production and supporting various anabolic pathways that produce the macromolecules required for cell growth and proliferation. Flux through the different metabolic pathways can be modulated in response to varying growth conditions and demands, as seen in cancer cells, which adopt high rates of aerobic glycolysis that is characterized by extensive glucose uptake and high levels of lactate production (1). Glycolysis can be regulated by changes in the expression or activity of enzymes that catalyze various steps of the pathways. Allosteric regulation or covalent modification of many metabolic enzymes has been described, and the concentration of the active enzyme can also be controlled through changes in transcription, splicing, or protein stability. The subcellular localization of some glycolytic enzymes has also been shown to play an important role in regulating their functions (2, 3). In the first step of glucose metabolism, hexokinases (HK) generate glucose-6-phosphate, which can then be further used through glycolysis to produce energy, diverted into the pentose phosphate pathway (PPP) to produce NADPH and anabolic intermediates or converted to glycogen for storage. HK2 is thought to play a key role in promoting anabolic pathways and is frequently overexpressed in cancers (4). Both HK1 and HK2 show cytoplasmic and—through interaction with voltage-dependent anion-selective channel (VDAC)—mitochondrial localization. The presence of HK at the mitochondria provides a close proximity to intramitochondrial ATP production, which helps to couple glycolysis and oxidative phosphorylation (5). This mitochondrial HK also helps to limit mitochondrial reactive oxygen species (ROS) production by maintaining local ADP levels (6). Furthermore, mitochondrial HK2 can directly inhibit apoptosis through regulation of the mitochondrial permeability transition pore that, in part, reflects the ability to block the recruitment of proapoptotic proteins such as Bax and Bak (7–11). The interaction of HK2 with the phosphoprotein PEA15 is also required for protection from apoptosis under hypoxia (12). Several oncogenes and tumor suppressor genes can regulate the expression of different glycolytic enzymes (13, 14). The p53 tumor suppressor protein can drive the expression of a number of genes that result in an increased rate of oxidative phosphorylation while dampening glycolysis. In this way, p53 can oppose the acquisition of aerobic glycolysis that is characteristic of cancer cell metabolism (15). One of these p53 target genes www.pnas.org/cgi/doi/10.1073/pnas.1206530109

encodes for the protein TIGAR, which functions as a fructose2,6-bisphosphatase (Fru-2,6-BPase) (16, 17). TIGAR limits the levels of fructose-2,6-bisphosphate (Fru-2,6-BP) in the cell and subsequently lowers the activity of phosphofructosekinase-1 (PFK1)—a key step in the control of glycolysis (18). This reduction in glycolytic flux promotes the PPP, which generates NADPH. TIGAR expression is therefore able to increase NADPH levels and promote antioxidant function, thereby limiting ROS-associated apoptosis and autophagy (16, 19–22). However, the ability of TIGAR to promote cell survival is cell and context dependent. In cells that highly depend on the maintenance of glycolytic flux for survival, expression of TIGAR correlates with decreased, rather than increased, survival (16, 23). In this report, we show that TIGAR can form a complex with HK2 at the mitochondria and modulate HK2 activity under hypoxia, so contributing to the control of mitochondrial ROS and cell survival under hypoxia. Results Our previous work showed that TIGAR can function to limit ROS accumulation, autophagy, and apoptosis, and that this function of TIGAR is most clearly seen under conditions of metabolic stress that include hypoxia (19). Using A2780 and SW48 cells, each of which express clearly detectable TIGAR protein levels, we examined the effect of hypoxia on TIGAR levels and localization (Fig. 1). Culturing cells in 0.1% (vol/ vol) O2 for 24 h did not affect the overall levels of TIGAR expression, but cell fractionation revealed that hypoxia induced the relocalization of some TIGAR protein to the mitochondrial fraction (Fig. 1A). Direct staining of cells for TIGAR and Tom 20, a mitochondrial marker, confirmed that localization of TIGAR to the mitochondria was not observed in cells grown under 20% (vol/vol) O2, but was induced when cells were grown in 0.1% O2 (Fig. 1 B and C). Treatment of cells with DPCA or DMOG, small molecules that lead to the stabilization of HIF1α (24, 25), also promoted the relocalization of TIGAR to mitochondria (Fig. 1 D and E)—although it should be noted that these drugs generally inhibit prolyl 4-hydroxylases and will therefore have additional, HIF1α-independent effects. Specificity of the TIGAR staining in these immunofluorescence experiments was demonstrated by using previously described siRNAs to deplete TIGAR protein (Fig. S1A). To determine where in the mitochondria TIGAR is located, we performed immunofluorescence studies in hypoxic cells treated with a gradient of digitonin to permeabilize different mitochondrial membranes (26). Using Tom20 as a marker of the outer membrane, HtrA2 as a marker of the intermembrane space and Hsp60 as a marker of

Author contributions: E.C.C. and K.H.V. designed research; E.C.C. and R.L.L. performed research; E.C.C., R.L.L., and K.H.V. analyzed data; and E.C.C. and K.H.V. wrote the paper. The authors declare no conflict of interest. This article is a PNAS Direct Submission. Freely available online through the PNAS open access option. 1

To whom correspondence should be addressed. E-mail: [email protected].

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. 1073/pnas.1206530109/-/DCSupplemental.

PNAS | December 11, 2012 | vol. 109 | no. 50 | 20491–20496

CELL BIOLOGY

The p53-inducible protein TIGAR (Tp53-induced Glycolysis and Apoptosis Regulator) functions as a fructose-2,6-bisphosphatase (Fru-2,6-BPase), and through promotion of the pentose phosphate pathway, increases NADPH production to help limit reactive oxygen species (ROS). Here, we show that under hypoxia, a fraction of TIGAR protein relocalized to mitochondria and formed a complex with hexokinase 2 (HK2), resulting in an increase in HK2 activity. Mitochondrial localization of TIGAR depended on mitochondrial HK2 and hypoxia-inducible factor 1 (HIF1α) activity. The ability of TIGAR to function as a Fru-2,6-BPase was independent of HK2 binding and mitochondrial localization, although both of these activities can contribute to the full activity of TIGAR in limiting mitochondrial ROS levels and protecting from cell death.

Fig. 1. Relocalization of TIGAR to mitochondria during hypoxia. (A) Western blot analysis on whole cell lysate (Left) and on mitochondrial fraction (Right) of A2780 and SW48 cells after 24 h in control (normoxia) and 0.1% O2. mtHsp70 was used as a loading control for mitochondria. SW48 (B) and A2780 (C) cells were incubated in control and 0.1% O2. After 24 h, cells were fixed and stained with anti-TIGAR antibody and anti-Tom20 antibody to visualize mitochondria. (D) HeLa cells were treated with 100 μM DPCA. After 24 h, cells were fixed and stained with anti-TIGAR and anti-Tom20 antibody to visualize mitochondria. DMSO only was used as control. (E) A2780 cells were treated with 1 mM DMOG. After 24 h, cells were fixed and stained with anti-TIGAR and anti-Tom20 antibody to visualize mitochondria. (F) HeLa cells exposed to 1% O2 for 24 h were fixed and then permeabilized by using the indicated concentration of digitonin. Cells were then stained with the indicated antibody to visualize different mitochondrial compartments. (Scale bars: B and D, 40 μm; C, 10 μm; E, 20 μm.)

the matrix, we were able to show that TIGAR located with Tom20, at the outer mitochondrial membrane (Fig. 1F). To investigate the function of TIGAR at the outer mitochondrial membrane, we considered previous studies that identified the binding between the bisphosphatase domain of PFK2/ FBPase and glucokinase (27–29), a member of the HK family expressed mainly in the liver, pancreas, gut, and brain (5). Because TIGAR resembles the bisphosphatase domain of PFK2/ FBPase (16, 17), we reasoned that TIGAR might form a complex with HK family members. Of the four HK isoforms, HK2 is the most commonly overexpressed in cancers and cancer cell lines. Using A2780, SW48, and CACO2 cells, we looked for coprecipitation of endogenous HK2 with endogenous TIGAR under conditions of glucose depletion or hypoxia (0.1% O2) (Fig. 2A). An unrelated isotype-matched antibody was used as negative control (Fig. S1B). As shown previously, hypoxia induced an increase in HK2 expression (Fig. S1C) (30). Although no or low levels of binding between TIGAR and HK2 could be detected under normal growth conditions, hypoxia resulted in a clear increase in the association of HK2 with TIGAR in all three cell lines (Fig. 2A). This hypoxia-induced association of TIGAR with

HK2 was abrogated when the cells were also subject to glucose starvation. Immunofluorescence studies confirmed that the colocalization of TIGAR with HK2 under hypoxia was lost in the absence of glucose (Fig. 2B). Interestingly, the glucose dependence of mitochondrial localization has also been reported for HK2, which is localized to the mitochondria in the presence of glucose but moves to the cytosol when glucose is removed (2, 12). Consistently, we also found that the removal of glucose resulted in some dissociation of HK2 and a mitochondrial marker (mtHsp70), particularly in 0.1% O2, although even under these conditions, some HK2 remained at the mitochondria (Fig. 2C). Importantly, the HK2 binding and relocalization of TIGAR was also seen in other cell lines (U2OS, HCT116, and HeLa) and under 1% O2 (Fig. S1 D, E, and H). Finally, we were able to show that the mitochondrial relocalization of TIGAR in hypoxia depends on HK2, because cells depleted of HK2 (Fig. S1F) do not relocalize TIGAR in 0.1% O2 (Fig. 2D). However, overexpression of HK2 under normoxia is not sufficient to move TIGAR to the mitochondria (Fig. S1G), suggesting that other hypoxia-dependent events are necessary for the relocalization. Furthermore, we were unable to detect an interaction between TIGAR and HK1 or

Fig. 2. TIGAR interacts with HK2 during hypoxia. (A) Cell lysates from A2780, CaCO2, and SW48 cells with indicated treatments (hypoxia = 0.1% O2) were immunoprecipitated with anti-TIGAR antibody. The membranes were probed with the indicated antibodies. (B) HeLa cells grown in the indicated conditions (hypoxia = 0.1% O2) for 24 h were fixed and stained with the indicated antibodies. (C) HeLa cells grown in the indicated conditions (hypoxia = 0.1% O2) for 24 h were fixed and stained with anti-HK2 and anti-mtHsp70 to visualize mitochondria. (D) HeLa cells treated with control siRNA or HK2 siRNA that were grown in normoxia or 0.1% O2 for 24 h were fixed and stained with anti-TIGAR antibody and anti-Tom20 antibody to visualize mitochondria. (E) Cells transfected with the indicated plasmids or siRNA were grown in normoxia or hypoxia (0.1% O2) for 24 h and hexokinase (HK) activity measured. The relative activity was obtained by normalization to the cells with control siRNA. *P < 0.05 compared with control. (Scale bars: 20 μm.)

20492 | www.pnas.org/cgi/doi/10.1073/pnas.1206530109

Cheung et al.

Fig. 4. TIGAR maintains cell survival via HK2 during hypoxia. (A) Cells transfected with empty vector and control siRNA (control), wild-type TIGAR, HK2, or indicted siRNAs were grown in normoxia (open bar) or hypoxia (0.1% O2) (filled bars) for 24 h. (A) Mitochondrial ROS level. (B) Mitochondrial membrane potential was measured by TMRE fluorescence. (C) Cell death was measured by PI exclusion. *P < 0.05 compared with hypoxia control. (D) Western blot showing representative example of expression of TIGAR and HK2 protein in cells treated in A–C.

Fig. 3. TIGAR interaction with HK2 during hypoxia depends on HIF1α. (A) A2780 cells treated with the indicated compounds that were grown in 0.1% O 2 for 24 h were fixed and stained with anti-Tom20 to visualize mitochondria and anti-TIGAR to determine TIGAR location. (B) A2780 cells treated with either control siRNA or HIF1α siRNA were grown in hypoxia (0.1% O 2) for 24 h, then fixed and stained with anti-TIGAR and anti-Tom20 to visualize mitochondria. (C ) Cell lysates from A2780 cells treated with control siRNA or HIF1α siRNA and grown in normoxic or hypoxic conditions were immunoprecipitated with anti-TIGAR antibody. The membranes were probed with the indicated antibodies. (D) Cell lysates from A2780 cells treated with CAY10582 as indicated and grown in normoxic or hypoxic (0.1% O2) conditions were immunoprecipitated with anti-TIGAR antibody. The membranes were probed with the indicated antibodies. (Scale bars: 15 μm.)

TIGAR or HK2 increased ROS levels (Fig. 4A). Interestingly, depletion of HK2 abrogated the ability of TIGAR overexpression to limit the mitochondrial ROS, suggesting that HK2 is required for this function of TIGAR (Fig. 4A). HK2 activity has also been shown to help maintain mitochondrial membrane potential (9) and is consistent with an ability of TIGAR to enhance HK2 activity. The overexpression of TIGAR rescued loss of membrane potential, whereas depletion of TIGAR exacerbate the decrease of membrane potential compared with control cells in hypoxia (Fig. 4B). Elevated mitochondrial ROS levels can lead to the induction of apoptosis, and mitochondrial HK2 has been shown to protect cells from apoptosis through several mechanisms (7, 8, 37). We therefore examined the effect of TIGAR expression on cell survival during hypoxia. Consistent with the ability of TIGAR to maintain mitochondrial membrane potential and increase HK2 activity, increased TIGAR levels protected against cell death during hypoxia, whereas the depletion of TIGAR or HK2 enhanced cell death (Fig. 4C). HK2 protein levels were assessed by Western blot (Fig. 4D). TIGAR can function as a Fru-2,6-BPase, with a critical role for three amino acid residues within the region similar to the catalytic domain of the bisphosphatase (16). Substitutions of each of these amino acids in the mutant TIGAR TM (H11A/ E102A/H198A) (Fig. 5A) results in the loss of bisphosphatase activity. However, despite the loss of enzyme activity, this mutant TIGAR protein was still able to relocalize to mitochondria in response to hypoxia, in a manner similar to the wild-type protein (Fig. 5B). A screen of additional TIGAR mutants showed that a small deletion within the C terminus of the TIGAR protein (TIGAR 258–261), away from the catalytic domain, prevented hypoxia-induced relocalization to the mitochondria (Fig. 5B), while retaining the normal localization in normoxia (Fig. S2A). Despite deleting only four amino acids,

Cheung et al.

PNAS | December 11, 2012 | vol. 109 | no. 50 | 20493

CELL BIOLOGY

VDAC under hypoxia (Fig. S1H), and the levels of HK1 expression in these cells did not impact on the ability of HK2 to interact with TIGAR (Fig. S1I). Previous studies have suggested that binding to PFK2/FBPase enhances the activity of glucokinase (27, 31, 32), and so we examined the effect of TIGAR on HK activity under normoxic and hypoxic conditions. Although HK2 depletion by siRNA strongly reduced HK activity in both conditions, a clear reduction in HK activity was also seen after siRNA depletion of TIGAR (Fig. 2E) under hypoxia, although TIGAR depletion did not influence HK activity in 20% O2. Conversely, overexpression of TIGAR resulted in enhanced HK2 activity, which was again strongly inhibited by siRNA depletion of HK2 (Fig. 2E). These results show that hypoxia promotes the interaction of TIGAR with HK2, resulting in the mitochondrial localization of TIGAR and enhanced HK activity. The results shown in Fig. 1 suggest that TIGAR relocalizes to the mitochondria when HIF1α is activated. Using two inhibitors of HIF1α accumulation, compound c (which inhibits AMPK and HIF1α activation under hypoxia) (33, 34) and CAY10585 (which prevents the accumulation of HIF1α) (35), we found that the hypoxia-induced relocalization of TIGAR depended on HIF1α (Fig. 3A). siRNA-mediated depletion of HIF1α also abrogated the relocalization of TIGAR to mitochondria (Fig. 3B). Interestingly, depletion of HIF1α also prevented the binding of HK2 to TIGAR (Fig. 3C), as did inhibition of HIF1α activation by CAY10585 (Fig. 3D). We have found that the TIGAR can regulate ROS levels by promoting the PPP and the generation of reduced glutathione (16). However, mitochondrial HK activity has also been shown to inhibit mitochondrial ROS and promote cell survival (6, 9, 36). We therefore examined the effect of modulating TIGAR on these responses in hypoxic conditions, when TIGAR is associated with HK2. Under these conditions, overexpression of wild-type TIGAR significantly lowered mitochondrial ROS, whereas depletion of

Fig. 5. Relocalization of TIGAR mutants during hypoxia. (A) Schematic of the various TIGAR mutants. (B) HeLa cells (Left) and A2780 cells (Right) transfected with wild-type TIGAR or the indicated mutants were grown in hypoxia (1% O2) at 30 °C for 24 h. Cells were then fixed and stained with antiTIGAR and anti-Tom20 to visualize mitochondria. (Scale bar: 15 μm.) (C) Cell lysates from HeLa cells transfected with wild-type TIGAR or the indicated mutants and grown at 30 °C in hypoxia (1% O2) were immunoprecipitated by using anti-TIGAR antibody. The membranes were probed with the indicated antibodies. (D) Cells transfected with wildtype TIGAR or the indicated mutants were grown in normoxia (control) or hypoxia (1% O2) for 24 h at 30 °C and their HK activity measured. The relative activity was obtained by normalization to the control cells in normoxia. *P < 0.05 compared with hypoxia control. (E) Fru-2,6-BP level in cells transfected with the indicated TIGAR constructs (empty vector as control) that were exposed to hypoxia (1% O2) for 24 h. *P < 0.05 compared with control.

TIGAR 258–261 was very poorly expressed under hypoxia at 37 °C and comparable expression levels of all TIGAR proteins were achieved by growing the transfected cells at 30 °C. The mitochondrial localization of the TIGAR mutants was confirmed by cell fractionation and Western blotting (Fig. S2B). To determine whether mitochondrial relocalization correlated with HK2 binding, we carried out further coimmunoprecipitation studies with transfected TIGAR and endogenous HK2. Both wild-type and TM proteins retained the ability to bind HK2 in response to hypoxia, correlating with their ability to move to the mitochondria, whereas 258–261 or the double mutant TM/258–261 failed to bind (Fig. 5C). Similarly, these mutants failed to relocalise with HK2 during hypoxia compared with the wild-type and TM proteins (Fig. S2C). Having established that the 258–261 mutant cannot translocate to the mitochondria or bind to HK2, we determined the effects of this mutant on HK2 activity. As expected, under normoxic conditions, where no TIGAR/HK2 interaction was seen, TIGAR expression did not impact HK2 activity. In hypoxia, both wild-type and TM enhanced HK2 activity (Fig. 5D), indicating 20494 | www.pnas.org/cgi/doi/10.1073/pnas.1206530109

that the ability of TIGAR to regulate HK2 activity does not depend on its Fru-2,6-BPase function. However, the TIGAR mutants that did not bind HK2 (258–261 and TM/258–261) were unable to enhance HK2 activity (Fig. 5D). Finally, we showed that loss of the ability to bind HK2 did not impact the Fru-2,6-BPase function of TIGAR (Fig. 5E), because the 258–261 mutant retained a similar ability to decrease intracellular Fru-2,6-BP levels as wild-type TIGAR or the FBPase-2 domain of PFK2. We next sought to determine some of the effects of TIGAR on mitochondrial function. Mitochondrial ROS levels were extremely low in these cells under normoxia (Fig. 6A) and not significantly impacted by TIGAR expression. However, the limitation of mitochondrial ROS in hypoxic cells by TIGAR was also seen after expression of the TM mutant (Fig. 6A), indicating that the ability of TIGAR to promote HK2 function can help to limit ROS independently of Fru-2,6-BPase function. Neither of the HK2 binding deficient TIGAR mutants (258–261 and TM/ 258–261) showed effective mitochondrial ROS limiting activity (Fig. 6A). We noted, however, that the TM mutant showed a somewhat lower efficiency in limiting mitochondrial ROS Cheung et al.

compared with the wild-type protein, suggesting that Fru-2,6BPase function may also make some contribution to antioxidant activities under hypoxia (Fig. 6A). A similar activity of the TIGAR mutants in regulating the mitochondrial membrane potential was seen, where under hypoxia, TIGAR mutants that cannot bind to HK2 failed to protect against the decrease of mitochondrial membrane potential (Fig. 6B). Again, the TM mutant retained this activity (Fig. 6B). Taken together, these data suggest that in hypoxia, the HK2 binding activity of TIGAR plays an important role in lowering ROS levels and maintaining mitochondrial membrane potential—and that under these conditions, the Fru-2,6-BPase activity of TIGAR may contribute to, but is not necessary for this activity. Turning to an assessment of the role of TIGAR on cell death, we found that although both the TM and 258–261 mutants retain some ability to protect against cell death, albeit with diminished activity, this is lost in the TM/258–261 double mutant (Fig. 6C). Equal level of TIGAR protein expression was confirmed by Western blot (Fig. 6D). Together with the results shown in Fig. 4, these data suggest that the binding to HK2 is important for TIGAR to protect cells against hypoxic cell death, with a cooperative contribution of the Fru-2,6-BPase activity. Mitochondrial ROS has been shown to activate HIF1α (38, 39), although we were unable to see any defect in HIF1 activation under hypoxia after overexpression of TIGAR (Fig. S3A). Depletion of HIF1α results in enhanced death in hypoxic cells (40), and we found that the expression of TIGAR was unable to rescue the very high levels of cell death seen under these conditions (Fig. S3B). Because TIGAR cannot localize to the mitochondria in the absence of HIF1α, these results support the suggestion that mitochondrial localization and HK2 binding are important for the survival function of TIGAR in response to hypoxia. Cheung et al.

PNAS | December 11, 2012 | vol. 109 | no. 50 | 20495

CELL BIOLOGY

Fig. 6. TIGAR mutant that is unable to bind to HK2 fails to protect cells during hypoxia. Cells transfected with the indicted TIGAR mutants were grown in control or hypoxia (1% O 2) for 24 h at 30 °C. (A) Mitochondrial ROS level. (B) Mitochondrial membrane potential was measured by TMRE fluorescence. (C ) Cell death was measured by PI exclusion. *P < 0.05 compared with hypoxia control. (D) Western blot showing representative example of expression of TIGAR and HK2 protein in cells treated in A–C.

Discussion TIGAR has been shown to function as a Fru-2,6-BPase, regulating intracellular Fru-2,6-BP levels and glycolysis under various conditions (16, 17, 21, 23). In this study, we identified an additional function for TIGAR in lowering ROS and maintaining cell survival during hypoxia through a mechanism that is independent of Fru-2,6-BPase activity. This function is achieved by translocation of TIGAR to the mitochondria and interaction with HK2, resulting in enhanced HK2 activity, the regulation of mitochondrial membrane potential, and decreased mitochondrial ROS. Although hypoxia alone can increase glycolysis by increasing the expression of the glycolytic enzymes, including HK2 itself (41), our data indicate that mitochondrial HK2 activity can be further enhanced by interaction with TIGAR. Interestingly, the subcellular localization of HK2 has been shown to determine which pathway for glucose metabolism is used, with cytoplasmic HK2 promoting anabolic pathways such as the PPP or glycogen synthesis and mitochondrial HK2 directing glucose through glycolysis (2). The function of mitochondrial TIGAR may therefore be quite distinct from the ability to promote the PPP, although there will clearly be intersections of the consequences of these activities in terms of ROS regulation and apoptosis control. In addition to promoting glycolysis, several mechanisms through which mitochondrial HK2 can restrain apoptosis have been proposed. In general, overexpression of HK2 can lower the level of mitochondrial ROS, and this activity depends on the mitochondrial localization and glucose phosphorylation activity of HK2 (6, 9). This function of HK2 can protect cells against oxidative stress induced cell death (36). Mitochondrial HK has also been suggested to inhibit cell death by preventing the activation of the mitochondrial permeability transition (7–9) or by abrogating the activity of the proapoptotic Bcl-2 family proteins such as Bax and Bak (9–11). Consistent with these reports, we have also shown here that TIGAR can only decrease mitochondria ROS during hypoxia when HK2 is present on the mitochondria. Regulation of HK2 localization to the mitochondria has been suggested to act as a molecular switch that can sense the level of glucose and determine the fate of the cell. Although HK2 can protect cells from hypoxic cell death, this activity is lost in the absence of glucose (12), conditions under which HK2 becomes dissociated from the mitochondria (2). Consistently, we have also shown that the translocation of TIGAR to the mitochondria and the binding to HK2 depend on glucose availability. The direct signal that is responsible for the translocation of HK2 is not yet definitively shown, but it has been suggested that Aktdependent phosphorylation (37, 42) or HK2 conformational changes after binding to glucose (43) may play a role. The localization of TIGAR and HK2 at the mitochondria during hypoxia may therefore offer several mechanisms to integrate energy metabolism with cell survival, through the control of glucose metabolism, ROS level, and mitochondrial outer membrane permeabilization. Interestingly, although we show here that TIGAR mutants lacking Fru-2,6-BPase activity retain some ability to limit mitochondrial ROS and cell death in hypoxic cells, our previous studies showed that under conditions of both hypoxia and glucose depletion (where neither TIGAR nor HK2 are at the mitochondria), the protective effect of TIGAR depended entirely on Fru-2,6-BPase activity (19). These data suggest that TIGAR shows two independent functions: a Fru-2,6-BPase activity that is manifest in normoxic, hypoxic, and glucose limited cells and functions to promote the PPP, produce NADPH, and limit ROS, and a second hypoxia-induced activity that depends on HIF1α and glucose, and involves mitochondrial localization, binding, and activation of HK2. The interaction between HK2 and TIGAR is reminiscent of the interaction between glucokinase (a HK family member) and 6-phosphofructo-2-kinase/fructose-2,6-bisphosphatase (PFKFB) at the bisphosphatase domain (27–29). Similar to overexpression of TIGAR in our system, increasing PFKFB level can also

increase GK activity in pancreatic β-cells (27, 31, 32), an activity that depends on the availability of glucose in these cells (29). These results underline an important role of these interactions as sensors of glucose and oxygen availability to modulate energy metabolism and cell survival. TIGAR expression is up-regulated by the tumor suppressor p53 and is likely to contribute to the ability of p53 to help cells adapt to metabolic stress. However, the prosurvival and HK2 stimulating activities of TIGAR in hypoxic conditions may also contribute to the survival of cells during malignant progression, making it possible that deregulated expression of TIGAR could promote, rather than inhibit, cancer development. Materials and Methods Detailed methods are in SI Materials and Methods.

constructs were kindly provided by Philipp Mergenthaler (Berlin, Germany) (44). Commercially available and the described Anti-TIGAR antibody (16) were used for protein analysis. Immunofluorescence. Cell fixation and immunofluorescence were performed as described (16, 45). Visualization of proteins at different mitochondrial compartments by using a gradient of digitonin (Sigma) concentration to permeabilize membranes was performed as described (26). Mitochondrial Membrane Potential, Mitochondrial ROS, and Cell Death. Live mitochondrial membrane potential was determined as described (45). Hexokinase activity and Fru-2,6-BP Level. Fru-2,6-BP levels and HK activity was measured as described (16, 37).

Cell Lines, Transfections, siRNAs, and Antibodies. Cell lines were transfected by using Lipofectamine 2000 reagent (Invitrogen). siRNA and TIGAR constructs were described (Santa Cruz) (16) or generated for this study. HK2

ACKNOWLEDGMENTS. We thank Eyal Gottlieb for advice and critical reading of the manuscript; Andreas Hock and Patricia Muller for technical advice; and Philipp Mergenthaler for the HK2 expression plasmids. Work in the K.H.V. laboratory is supported by Cancer Research UK. E.C.C. is supported by a fellowship from the Canadian Institutes of Health Research.

1. Vander Heiden MG, Cantley LC, Thompson CB (2009) Understanding the Warburg effect: The metabolic requirements of cell proliferation. Science 324(5930):1029–1033. 2. John S, Weiss JN, Ribalet B (2011) Subcellular localization of hexokinases I and II directs the metabolic fate of glucose. PLoS ONE 6(3):e17674. 3. Sirover MA (2011) On the functional diversity of glyceraldehyde-3-phosphate dehydrogenase: Biochemical mechanisms and regulatory control. Biochim Biophys Acta 1810(8):741–751. 4. Mathupala SP, Ko YH, Pedersen PL (2006) Hexokinase II: Cancer’s double-edged sword acting as both facilitator and gatekeeper of malignancy when bound to mitochondria. Oncogene 25(34):4777–4786. 5. Wilson JE (2003) Isozymes of mammalian hexokinase: Structure, subcellular localization and metabolic function. J Exp Biol 206(Pt 12):2049–2057. 6. da-Silva WS, et al. (2004) Mitochondrial bound hexokinase activity as a preventive antioxidant defense: Steady-state ADP formation as a regulatory mechanism of membrane potential and reactive oxygen species generation in mitochondria. J Biol Chem 279(38):39846–39855. 7. Azoulay-Zohar H, Israelson A, Abu-Hamad S, Shoshan-Barmatz V (2004) In self-defence: Hexokinase promotes voltage-dependent anion channel closure and prevents mitochondria-mediated apoptotic cell death. Biochem J 377(Pt 2):347–355. 8. Chiara F, et al. (2008) Hexokinase II detachment from mitochondria triggers apoptosis through the permeability transition pore independent of voltage-dependent anion channels. PLoS One 3(3):e1852. 9. Sun L, Shukair S, Naik TJ, Moazed F, Ardehali H (2008) Glucose phosphorylation and mitochondrial binding are required for the protective effects of hexokinases I and II. Mol Cell Biol 28(3):1007–1017. 10. Majewski N, Nogueira V, Robey RB, Hay N (2004) Akt inhibits apoptosis downstream of BID cleavage via a glucose-dependent mechanism involving mitochondrial hexokinases. Mol Cell Biol 24(2):730–740. 11. Pastorino JG, Shulga N, Hoek JB (2002) Mitochondrial binding of hexokinase II inhibits Bax-induced cytochrome c release and apoptosis. J Biol Chem 277(9):7610–7618. 12. Mergenthaler P, et al. (2012) Mitochondrial hexokinase II (HKII) and phosphoprotein enriched in astrocytes (PEA15) form a molecular switch governing cellular fate depending on the metabolic state. Proc Natl Acad Sci USA 109(5):1518–1523. 13. Levine AJ, Puzio-Kuter AM (2010) The control of the metabolic switch in cancers by oncogenes and tumor suppressor genes. Science 330(6009):1340–1344. 14. Jones RG, Thompson CB (2009) Tumor suppressors and cell metabolism: A recipe for cancer growth. Genes Dev 23(5):537–548. 15. Vousden KH, Ryan KM (2009) p53 and metabolism. Nat Rev Cancer 9(10):691–700. 16. Bensaad K, et al. (2006) TIGAR, a p53-inducible regulator of glycolysis and apoptosis. Cell 126(1):107–120. 17. Li H, Jogl G (2009) Structural and biochemical studies of TIGAR (TP53-induced glycolysis and apoptosis regulator). J Biol Chem 284(3):1748–1754. 18. Albert L, Lehninger DLN, Cox MM (1995) Principles of Biochemistry. (Worth Publishers, New York), 2nd Ed. 19. Bensaad K, Cheung EC, Vousden KH (2009) Modulation of intracellular ROS levels by TIGAR controls autophagy. EMBO J 28(19):3015–3026. 20. Yin L, Kosugi M, Kufe D (2012) Inhibition of the MUC1-C oncoprotein induces multiple myeloma cell death by down-regulating TIGAR expression and depleting NADPH. Blood 119(3):810–816. 21. Peña-Rico MA, et al. (2011) TP53 induced glycolysis and apoptosis regulator (TIGAR) knockdown results in radiosensitization of glioma cells. Radiother Oncol 101(1): 132–139. 22. Lui VW, et al. (2011) Inhibition of c-Met downregulates TIGAR expression and reduces NADPH production leading to cell death. Oncogene 30(9):1127–1134. 23. Kimata M, et al. (2010) p53 and TIGAR regulate cardiac myocyte energy homeostasis under hypoxic stress. Am J Physiol Heart Circ Physiol 299(6):H1908–H1916. 24. Franklin TJ, Morris WP, Edwards PN, Large MS, Stephenson R (2001) Inhibition of prolyl 4-hydroxylase in vitro and in vivo by members of a novel series of phenanthrolinones. Biochem J 353(Pt 2):333–338.

25. Baader E, Tschank G, Baringhaus KH, Burghard H, Günzler V (1994) Inhibition of prolyl 4-hydroxylase by oxalyl amino acid derivatives in vitro, in isolated microsomes and in embryonic chicken tissues. Biochem J 300(Pt 2):525–530. 26. Otera H, Ohsakaya S, Nagaura Z, Ishihara N, Mihara K (2005) Export of mitochondrial AIF in response to proapoptotic stimuli depends on processing at the intermembrane space. EMBO J 24(7):1375–1386. 27. Smith WE, Langer S, Wu C, Baltrusch S, Okar DA (2007) Molecular coordination of hepatic glucose metabolism by the 6-phosphofructo-2-kinase/fructose-2,6- bisphosphatase:glucokinase complex. Mol Endocrinol 21(6):1478–1487. 28. Baltrusch S, Lenzen S, Okar DA, Lange AJ, Tiedge M (2001) Characterization of glucokinase-binding protein epitopes by a phage-displayed peptide library. Identification of 6-phosphofructo-2-kinase/fructose-2,6-bisphosphatase as a novel interaction partner. J Biol Chem 276(47):43915–43923. 29. Langer S, Kaminski MT, Lenzen S, Baltrusch S (2010) Endogenous activation of glucokinase by 6-phosphofructo-2-kinase/fructose-2,6-bisphosphatase is glucose dependent. Mol Endocrinol 24(10):1988–1997. 30. Mathupala SP, Rempel A, Pedersen PL (2001) Glucose catabolism in cancer cells: Identification and characterization of a marked activation response of the type II hexokinase gene to hypoxic conditions. J Biol Chem 276(46):43407–43412. 31. Massa L, et al. (2004) Interaction of 6-phosphofructo-2-kinase/fructose-2,6-bisphosphatase (PFK-2/FBPase-2) with glucokinase activates glucose phosphorylation and glucose metabolism in insulin-producing cells. Diabetes 53(4):1020–1029. 32. Baltrusch S, Langer S, Massa L, Tiedge M, Lenzen S (2006) Improved metabolic stimulus for glucose-induced insulin secretion through GK and PFK-2/FBPase-2 coexpression in insulin-producing RINm5F cells. Endocrinology 147(12):5768–5776. 33. Chua YL, Hagen T (2011) Compound C prevents Hypoxia-Inducible Factor-1α protein stabilization by regulating the cellular oxygen availability via interaction with Mitochondrial Complex I. BMC Res Notes 4:117. 34. Emerling BM, Viollet B, Tormos KV, Chandel NS (2007) Compound C inhibits hypoxic activation of HIF-1 independent of AMPK. FEBS Lett 581(29):5727–5731. 35. Lee K, et al. (2007) (Aryloxyacetylamino)benzoic acid analogues: A new class of hypoxia-inducible factor-1 inhibitors. J Med Chem 50(7):1675–1684. 36. Ahmad A, et al. (2002) Elevated expression of hexokinase II protects human lung epithelial-like A549 cells against oxidative injury. Am J Physiol Lung Cell Mol Physiol 283(3):L573–L584. 37. Gottlob K, et al. (2001) Inhibition of early apoptotic events by Akt/PKB is dependent on the first committed step of glycolysis and mitochondrial hexokinase. Genes Dev 15(11):1406–1418. 38. Mansfield KD, et al. (2005) Mitochondrial dysfunction resulting from loss of cytochrome c impairs cellular oxygen sensing and hypoxic HIF-alpha activation. Cell Metab 1(6):393–399. 39. Guzy RD, et al. (2005) Mitochondrial complex III is required for hypoxia-induced ROS production and cellular oxygen sensing. Cell Metab 1(6):401–408. 40. Semenza GL (2011) Hypoxia-inducible factor 1: Regulator of mitochondrial metabolism and mediator of ischemic preconditioning. Biochim Biophys Acta 1813(7):1263–1268. 41. Riddle SR, et al. (2000) Hypoxia induces hexokinase II gene expression in human lung cell line A549. Am J Physiol Lung Cell Mol Physiol 278(2):L407–L416. 42. Miyamoto S, Murphy AN, Brown JH (2008) Akt mediates mitochondrial protection in cardiomyocytes through phosphorylation of mitochondrial hexokinase-II. Cell Death Differ 15(3):521–529. 43. Jurczak MJ, et al. (2007) Transgenic overexpression of protein targeting to glycogen markedly increases adipocytic glycogen storage in mice. Am J Physiol Endocrinol Metab 292(3):E952–E963. 44. Mergenthaler P, et al. (2012) Mitochondrial hexokinase II (HKII) and phosphoprotein enriched in astrocytes (PEA15) form a molecular switch governing cellular fate depending on the metabolic state. Proc Natl Acad Sci USA 109(5):1518–1523. 45. Cheung EC, et al. (2006) Dissociating the dual roles of apoptosis-inducing factor in maintaining mitochondrial structure and apoptosis. EMBO J 25(17):4061–4073.

20496 | www.pnas.org/cgi/doi/10.1073/pnas.1206530109

Cheung et al.