Cancer Cell
`
`Previews
`
`IDH1 Mutations in Gliomas:
`When an Enzyme Loses Its Grip
`
`Christian Frezza,1 Daniel A. Tennant,1 and Eyal Gottlieb1,*
`1Cancer Research UK, The Beatson Institute for Cancer Research, Glasgow G61 1BD, UK
`*Correspondence: e.gottlieb@beatson.gla.ac.uk
`DOI 10.1016/j.ccr.2009.12.031
`
`The growing interest in cancer metabolism is best demonstrated by the rapid progress made in studying
`isocitrate dehydrogenase (IDH) mutations since their discovery just over a year ago. In a recent study
`published in Nature, Dang et al. identified 2-hydroxyglutarate as a product of tumor-associated IDH mutants
`with potential oncogenic activities.
`
`is now almost a century since the
`It
`studies that first associated cellular meta-
`bolic changes with cancer. However, the
`recognition of a causal connection be-
`tween metabolic alterations and cancer
`formation was revealed only this decade.
`Ironically,
`it was genetics, rather than
`biochemistry,
`that enabled this break-
`through when genes encoding mitochon-
`drial enzymes of the tricarboxylic acid
`(TCA) cycle, succinate dehydrogenase
`(SDH) and fumarate hydratase (FH), were
`identified as bona fide tumor suppressors
`(King et al., 2006). Over the past year, new
`genetic studies placed another metabolic
`enzyme, isocitrate dehydrogenase (IDH),
`in the spotlight of cancer biology (Yan
`et al., 2009a). High-throughput sequenc-
`ing revealed that two of the three isoforms
`of IDH (IDH1 and IDH2) are mutated in
`high proportions in gliomas (Parsons
`et al., 2008; Yan et al., 2009b). However,
`unlike SDH and FH, IDH mutations do
`not follow Knudson’s two-hit model of
`tumor suppressor genes.
`In the new
`study, Dang et al. (2009) demonstrated
`that although IDH1 mutants lose their
`normal enzymatic activity in tumors, they
`gain a new one, generating a new pro-
`duct, 2-hydroxyglutarate, with potentially
`tumor-supporting actions (making it an
`onco-metabolite).
`Eukaryotic cells contain two classes of
`IDH enzymes according to dependence
`on either NAD+ or NADP+. These enzymes
`normally convert isocitrate to a-ketogluta-
`rate (aka 2-oxoglutarate), with the concur-
`rent reduction of NAD(P)+ to NAD(P)H
`(Figure 1). The two NADP+-dependent
`forms, IDH1 and IDH2, are cytosolic and
`mitochondrial,
`respectively.
`IDH3,
`the
`only NAD+-dependent IDH, is located at
`
`the mitochondria and is part of the TCA
`cycle. Rapid cycling of metabolites be-
`tween cytosol and mitochondria is a
`common feature of cellular metabolism.
`Metabolites entering the mitochondria
`can be processed for energy generation
`usually through the production of NADH
`in the TCA cycle whereas metabolites
`exported back to the cytosol take part in
`anabolic processes. The transport of
`metabolites is also coupled to electron
`exchange between mitochondrial and
`cytosolic NADH and NADPH, both of
`which cannot move across the mitochon-
`drial inner membrane (Figure 1). Because
`mitochondrial NADH operates in energy
`metabolism and cytosolic NADPH func-
`tions in anabolic processes and redox
`control,
`it
`is reasonable to expect
`changes in one or all of these processes
`in tumors carrying an IDH mutation.
`Until now, only mutations in IDH1 and 2
`were found in cancers, therefore leaving
`the TCA cycle untouched (Yan et al.,
`2009a). IDH1 mutations form the lion’s
`share of IDH mutations found in cancer,
`with IDH2 mutation being much less
`common. So far, gliomas have been
`shown as the cancer type most likely to
`contain IDH mutations.
`Interestingly,
`they seem to arise early in the develop-
`ment of a glioma, suggesting that it con-
`fers advantage early on in tumor progres-
`sion. One of the most striking features of
`IDH1 and 2 mutations is that it is always
`the same residue that is mutated: R132
`in IDH1 and R172 in IDH2. These residues
`create the hydrophilic interactions that
`allow the binding of isocitrate (Xu et al.,
`2004). The residues that are substituted
`for arginine are wide ranging, which
`strongly suggests that it is not the new
`
`the
`the replacement of
`residue, but
`arginine, which supports tumorigenesis
`by impairing isocitrate binding. Indeed,
`loss of
`IDH function was reported for
`these mutants and therefore IDH was
`suggested to be a tumor suppressor
`(Zhao et al., 2009). However, the fact
`that mutations were observed only on
`specific arginine residues and only on
`one allele of IDH1/2 with the other remain-
`ing wild-type (WT) led to the hypothesis
`that these are, in fact, gain- rather than
`loss-of-function mutations with onco-
`genic potential.
`The new work (Dang et al., 2009) started
`with large-scale metabolite quantification
`(metabolomics) of cells expressing either
`WT or
`tumor-derived mutant of
`IDH1
`(R132H). Only one significant metabolic
`change was observed in mutant-IDH1-
`expressing cells, which was a large
`accumulation of 2-hydroxyglutarate, a
`reduced form of a-ketoglutarate (Figure 1).
`Indeed, Dang et al. confirmed that the
`carbon backbone of
`the accumulated
`2-hydroxyglutarate is derived from gluta-
`mine, the major source of a-ketoglutarate
`in these cells (Figure 1). These results
`suggest that the mutant IDH1 changed
`its substrate specificity and directionality.
`In vitro enzymatic analysis confirmed this;
`whereas WT IDH1 converted isocitrate to
`a-ketoglutarate, several
`tumor-associ-
`ated mutants of IDH1 could no longer
`catalyze this reaction and instead reduced
`a-ketoglutarate to 2-hydroxyglutarate (but
`not to isocitrate). Structural comparison of
`the mutant and WT IDH1 revealed that
`mutations in R132 change the orientation
`of the catalytic site so the enzyme binds
`NADPH with higher affinity, a feature
`that supports reductase rather
`than
`
`Cancer Cell 17, January 19, 2010 ª2010 Elsevier Inc. 7
`
`Rigel Exhibit 1052
`Page 1 of 3
`
`

`

`Cancer Cell
`
`Previews
`
`cell-permeable a-ketoglutarate esters
`prevent HIF activation in cells expressing
`mutant IDH1 (Zhao et al., 2009) supports
`this model.
`The normal metabolic role of 2-hydrox-
`yglutarate is not completely understood
`but 2-hydroxyglutarate is not unnatural
`to cells.
`It can be generated by spe-
`cific a-ketoglutarate reductase enzymes
`(Struys, 2006) and oxidized back to
`a-ketoglutarate by 2-hydroxyglutarate
`dehydrogenases (2HGD) (Figure 1). The
`picture is further complicated by the exis-
`tence of two enantiomers of 2-hydroxy-
`glutarate with specific 2HGD for each.
`Mutations in 2HGD cause pathological
`accumulation of 2-hydroxyglutarate with
`different clinical features based on the
`enantiomer involved. Pathological accu-
`mulation of
`the L-2-hydroxyglutarate
`enantiomer is characterized by progres-
`sive neuronal defects and was recently
`linked to increased risk of brain tumors
`including gliomas (Aghili et al., 2009).
`This is strong support for the potential
`oncogenic role of 2-hydroxyglutarate,
`but with one caveat: Dang et al. demon-
`strated that mutant
`IDH1 generates
`D-2-hydroxyglutarate and not the L enan-
`tiomer. Accumulation of D-2-hydroxyglu-
`tarate is observed in D-2HGD-deficient
`patients and is associated with encepha-
`lopathy, cardiomyopathy, and more—
`but, so far, not with tumors (Struys,
`2006). It is possible that D-2-hydroxyglu-
`tarate, when reaching very high levels, is
`too toxic to have tumorigenic potential.
`This could have therapeutic significance
`because it may suggest that a small and
`transient pharmacological
`inhibition of
`2HGD, by raising the levels of 2-hydroxy-
`glutarate from protumorigenic to toxic,
`could specifically kill gliomas with IDH1
`mutations.
`
`REFERENCES
`
`Aghili, M., Zahedi, F., and Rafiee, E.
`J. Neurooncol. 91, 233–236.
`
`(2009).
`
`Dang, L., White, D.W., Gross, S., Bennett, B.D.,
`Bittinger, M.A., Driggers, E.M., Fantin, V.R., Jang,
`H.G., Jin, S., Keenan, M.C., et al. (2009). Nature
`462, 739–744.
`
`King, A., Selak, M.A., and Gottlieb, E.
`Oncogene 25, 4675–4682.
`
`(2006).
`
`MacKenzie, E.D., Selak, M.A., Tennant, D.A.,
`Payne, L.J., Crosby, S., Frederiksen, C.M., Wat-
`son, D.G., and Gottlieb, E. (2007). Mol. Cell. Biol.
`27, 3282–3289.
`
`Figure 1. The Roles of IDH Enzymes in the Exchange of Metabolites between
`the Mitochondria and the Cytosol, and Their Potential Role in Tumorigenesis
`Many isoforms of TCA cycle enzymes (light blue) also operate in the cytosol. They are important for
`synchronizing bioenergetic and anabolic needs by directing TCA cycle metabolites and electrons, in
`the forms of NAD(P)H, in and out the mitochondria (red). The two major carbon sources for these metab-
`olites are glucose and glutamine, which are catabolized via glycolysis (green) and glutaminolysis (purple),
`respectively. The three IDH isoenzymes are important players in these processes. IDH3 is part of the TCA
`cycle where it generates NADH as a fuel for energy production while IDH1 and 2 are important for shuttling
`electrons between the mitochondria and the cytosol. Although mutations in IDH1 are expected to hinder
`these processes, newly described work (Dang et al., 2009) proposes a new gain-of-function role for
`glioma-associated mutants of IDH1. R132 mutations of IDH1 generate a new enzyme with a-ketoglutarate
`reductase activity that produces 2-hydroxyglutarate and increased 2-hydroxyglutarate strongly correlates
`with cancer formation. But the tumorigenic mechanism is not yet understood. One possibility may be that
`2-hydroxyglutarate inhibits PHD activity by competing with a-ketoglutarate binding.
`Solid or dashed lines indicate direct or indirect metabolic links, respectively.
`
`oxidase activity. Furthermore, modeling
`a-ketoglutarate
`into
`the
`structure
`suggests a new orientation of the binding
`to a-ketoglutarate that can explain the
`formation of a new product, rather than
`simply running the reaction in reverse.
`Finally, Dang et al. demonstrated that
`2-hydroxyglutarate levels are 100-fold
`higher in human gliomas that carry R132
`mutations of IDH1 than in tumors with
`WT IDH1.
`These results revealed a new gain-of-
`function activity of
`the tumor-derived
`IDH1 mutants and strongly correlated
`the levels of 2-hydroxyglutarate with
`tumorigenesis. However, does this grant
`2-hydroxyglutarate the title ‘‘onco-metab-
`olite’’ as Dang et al. proposed? What
`might be these oncogenic functions of
`2-hydroxyglutarate?
`
`The loss of activity of two other TCA
`cycle enzymes mentioned earlier, SDH
`or FH, supports tumor
`formation by
`increasing the levels of their respective
`TCA cycle substrates, succinate or fuma-
`rate. These substrates inhibit the oxygen-
`sensing
`enzymes
`hypoxia-inducible
`factor prolyl hydroxylases (PHDs) by com-
`peting with their cosubstrate a-ketogluta-
`rate (MacKenzie et al., 2007). PHD inhibi-
`tion leads to the activation of the HIF
`transcription factor among other,
`less
`characterized, effects (King et al., 2006).
`It was previously demonstrated that
`PHDs are inhibited in cells carrying
`mutant IDH1 (Zhao et al., 2009). There-
`fore, it is possible that like succinate and
`fumarate,
`2-hydroxyglutarate
`inhibits
`PHD activity by competing with a-keto-
`glutarate (Figure 1). The observation that
`
`8 Cancer Cell 17, January 19, 2010 ª2010 Elsevier Inc.
`
`Rigel Exhibit 1052
`Page 2 of 3
`
`

`

`Cancer Cell
`
`Previews
`
`Parsons, D.W., Jones, S., Zhang, X., Lin, J.C.,
`Leary, R.J., Angenendt, P., Mankoo, P., Carter,
`H., Siu, I.M., Gallia, G.L., et al. (2008). Science
`321, 1807–1812.
`
`Xu, X., Zhao, J., Xu, Z., Peng, B., Huang, Q.,
`Arnold, E., and Ding, J. (2004). J. Biol. Chem.
`279, 33946–33957.
`
`Yan, H., Bigner, D.D., Velculescu, V., and Parsons,
`D.W. (2009a). Cancer Res. 69, 9157–9159.
`
`Struys, E.A.
`21–29.
`
`(2006). J.
`
`Inherit. Metab. Dis. 29,
`
`Yan, H., Parsons, D.W., Jin, G., McLendon, R.,
`Rasheed, B.A., Yuan, W., Kos, I., Batinic-Haberle,
`
`I., Jones, S., Riggins, G.J., et al. (2009b). N. Engl.
`J. Med. 360, 765–773.
`
`Zhao, S., Lin, Y., Xu, W., Jiang, W., Zha, Z., Wang,
`P., Yu, W., Li, Z., Gong, L., Peng, Y., et al. (2009).
`Science 324, 261–265.
`
`SUMO Boosts the DNA Damage
`Response Barrier against Cancer
`
`Jiri Bartek1,2,* and Zdenek Hodny2
`1Institute of Cancer Biology and Centre for Genotoxic Stress Research, Danish Cancer Society, Strandboulevarden 49, DK-2100
`Copenhagen, Denmark
`2Department of Genome Integrity, Institute of Molecular Genetics, ASCR v.v.i., Videnska 1083, CZ-142 20 Prague 4, Czech Republic
`*Correspondence: jb@cancer.dk
`DOI 10.1016/j.ccr.2009.12.030
`
`Cells exposed to genotoxic insults such as ionizing radiation activate a signaling cascade to repair the
`damaged DNA. Two recent articles published in Nature show that such genome maintenance requires modi-
`fications of tumor suppressor proteins BRCA1 and 53BP1 by the small ubiquitin-like modifier SUMO.
`
`Proper genome maintenance, ensured by
`the cellular DNA damage response (DDR)
`machinery, is a prerequisite for normal
`development and prevention of premature
`aging and diverse devastating diseases
`including cancer (Jackson and Bartek,
`2009). Indeed, one reason for cancer inci-
`dence not being even higher appears to be
`the intrinsic ability of our cells to detect
`and deal with the DNA damage caused
`by exogenous genotoxic agents such as
`radiation or chemicals as well as endoge-
`nous sources such as oncogene-evoked
`replication stress and telomere erosion
`during the early stages of cancer develop-
`ment (Halazonetis et al., 2008; Jackson
`and Bartek, 2009). Even if some DNA
`lesions, such as subsets of DNA double-
`strand breaks (DSB) that occur commonly
`during tumorigenesis, remain unrepaired,
`sustained signaling and effector path-
`ways within the DDR ‘‘anticancer barrier’’
`machinery usually eliminate such haz-
`ardous, genetically unstable cells by
`inducing cell death or a permanent cell
`cycle arrest known as cellular senescence
`(Halazonetis et al., 2008).
`From the mechanistic viewpoint, sens-
`ing, signaling, and repair of DSBs involve
`
`a plethora of proteins whose sequential
`accrual and function at the DNA damage
`sites is modulated by a myriad of post-
`translational modifications,
`including
`phosphorylation, acetylation, methyla-
`tion, and ubiquitylation, which are highly
`dynamic and reversible. The phosphoryla-
`tion/dephosphorylation events are per-
`formed by kinases such as the ATM, ATR,
`and DNA-PK, and several protein phos-
`phatases (Jackson and Bartek, 2009). The
`emerging ubiquitylation cascade com-
`prises the E3 ubiquitin ligases RNF8,
`RNF168, and BRCA1, as well as the E2
`ubiquitin-conjugating enzyme UBC13 and
`the candidate assembly factor HERC2
`(Bergink and Jentsch, 2009; Bekker-Jen-
`sen et al., 2010). Unlike the classical
`role of ubiquitylation in triggering protein
`degradation, however, this ubiquitin-medi-
`ated pathway orchestrates protein-protein
`interactions on damaged chromosomes
`and recruitment of the key DNA repair fac-
`tors 53BP1 and BRCA1 to DSBs, thereby
`promoting genomic integrity (Figure 1).
`Despite the rapid progress in under-
`standing the molecular basis of DSB
`signaling and repair, more surprises are
`in store for us in this lively area of
`
`research, as illustrated by two recent
`reports in Nature (Galanty et al., 2009;
`Morris et al., 2009). These exciting studies
`provide evidence for a key role of yet
`another protein modification, sumoylation
`(covalent attachment of the small proteins
`known as SUMO1, SUMO2, and SUMO3),
`in coordinating the DNA damage re-
`sponse to DSBs (Figure 1). Processes
`critical for cell fate decisions including
`survival and some aspects of DNA repair
`have been linked to the sumoylation
`pathway, particularly in yeast
`(Bergink
`and Jentsch, 2009; Branzei and Foiani,
`2008; Hay, 2005). However, the involve-
`ment of the sumoylation pathway in DSB
`response and its functional interplay with
`the ubiquitylation cascade that controls
`recruitment of 53BP1 and BRCA1 are
`novel and very relevant for genome main-
`tenance and protection against cancer.
`So what is revealed by the two new
`studies? First, in a complementary series
`of immunofluorescence and live-cell imag-
`ing experiments,
`they show that
`the
`SUMO1 and SUMO2/3 conjugates, as
`well as the E1 (SAE1), E2 (UBC9), and E3
`(PIAS1 and PIAS4) sumoylation enzymes,
`all rapidly accumulate at the sites of DNA
`
`Cancer Cell 17, January 19, 2010 ª2010 Elsevier Inc. 9
`
`Rigel Exhibit 1052
`Page 3 of 3
`
`