Correction

BIOCHEMISTRY Correction for “Identification of a small molecule inhibitor of 3-phosphoglycerate dehydrogenase to target serine biosynthesis in cancers,” by Edouard Mullarky, Natasha C. Lucki, Reza Beheshti Zavareh, Justin L. Anglin, Ana P. Gomes, Brandon N. Nicolay, Jenny C. Y. Wong, Stefan Christen, Hidenori Takahashi, Pradeep K. Singh, John Blenis, J. David Warren, Sarah-Maria Fendt, John M. Asara,GinaM.DeNicola,CostasA.Lyssiotis,LukeL.Lairson,and Lewis C. Cantley, which appeared in issue 7, February 16, 2016, of Proc Natl Acad Sci USA (113:1778–1783; first published February 1, 2016; 10.1073/pnas.1521548113). The editors note that the date on which this manuscript was sent for review was originally published incorrectly as December 7, 2015. The date should instead appear as November 2, 2015.

www.pnas.org/cgi/doi/10.1073/pnas.1602228113 CORRECTION

www.pnas.org PNAS | March 15, 2016 | vol. 113 | no. 11 | E1585 Downloaded by guest on September 30, 2021 Identification of a small molecule inhibitor of 3-phosphoglycerate dehydrogenase to target serine biosynthesis in cancers

Edouard Mullarkya,b,c, Natasha C. Luckid, Reza Beheshti Zavarehd, Justin L. Anglind, Ana P. Gomesa,e, Brandon N. Nicolayf, Jenny C. Y. Wonga,b, Stefan Christeng,h, Hidenori Takahashii,j,1, Pradeep K. Singhk,l, John Blenisa,e, J. David Warrenk,l, Sarah-Maria Fendtg,h, John M. Asaraj, Gina M. DeNicolaa,b, Costas A. Lyssiotism,n,2, Luke L. Lairsond,o,2, and Lewis C. Cantleya,b,2 aMeyer Cancer Center, Weill Cornell Medical College, New York, NY 10065; bDepartment of Medicine, Weill Cornell Medical College, New York, NY 10065; cBiological and Biomedical Sciences Graduate Program, Harvard Medical School, Boston, MA 02115; dThe California Institute for Biomedical Research, La Jolla, CA 92037; eDepartment of Pharmacology, Weill Cornell Medical College, New York, NY 10065; fMassachusetts General Hospital Cancer Center, Harvard Medical School, Charlestown, MA 02129; gLaboratory of Cellular Metabolism and Metabolic Regulation, Department of Oncology, Katholieke Universiteit Leuven, 3000 Leuven, Belgium; hLaboratory of Cellular Metabolism and Metabolic Regulation, Vesalius Research Center, Vlaams Instituut voor Biotechnologie Leuven, 3000 Leuven, Belgium; iDepartment of Systems Biology, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, MA 02115; jDivision of Signal Transduction, Department of Medicine, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, MA 02115; kDepartment of Biochemistry, Weill Cornell Medical College, New York, NY 10065; lMilstein Chemistry Core Facility, Weill Cornell Medical College, New York, NY 10065; mDepartment of Molecular and Integrative Physiology, University of Michigan, Ann Arbor, MI 48109; nDivision of Gastroenterology, Department of Internal Medicine, University of Michigan, Ann Arbor, MI 48109; and oDepartment of Chemistry, The Scripps Research Institute, La Jolla, CA 92037

Contributed by Lewis C. Cantley, December 23, 2015 (sent for review December 7, 2015; reviewed by Eyal Gottlieb and Brent R. Stockwell) Cancer cells reprogram their metabolism to promote growth and cancers (NSCLCs) overexpress PHGDH (15). Proliferation of proliferation. The genetic evidence pointing to the importance of the PHGDH-amplified cancer cell lines, and other lines that over- serine in tumorigenesis is striking. The encoding the express PHGDH without amplification, is inhibited by PHGDH 3-phosphoglycerate dehydrogenase (PHGDH), which cata- knockdown. In contrast, lines that express little PHGDH are re- BIOCHEMISTRY lyzes the first committed step of serine biosynthesis, is overexpressed sistant to shRNA-mediated ablation of the pathway, presumably in tumors and cancer cell lines via focal amplification and nuclear because serine import suffices (13, 14). A detailed mechanistic factor erythroid-2-related factor 2 (NRF2)-mediated up-regulation. understanding of why some cancer cells are addicted to serine PHGDH-overexpressing cells are exquisitely sensitive to genetic ab- synthesis despite the availability of extracellular serine for import lation of the pathway. Here, we report the discovery of a selective remains unclear. Interestingly, in triple negative breast cancer small molecule inhibitor of PHGDH, CBR-5884, identified by (TNBC) and NSCLC, PHGDH amplification and overexpression screening a library of 800,000 drug-like compounds. CBR-5884 are associated with more aggressive disease (13–16). Thus, PHGDH inhibited de novo serine synthesis in cancer cells and was selec- tively toxic to cancer cell lines with high serine biosynthetic activity. Significance Biochemical characterization of the inhibitor revealed that it was a noncompetitive inhibitor that showed a time-dependent onset of inhibition and disrupted the oligomerization state of PHGDH. The Serine supports a number of anabolic processes, including pro- identification of a small molecule inhibitor of PHGDH not only en- tein, , and nucleic acid synthesis. Cells can either import ables thorough preclinical evaluation of PHGDH as a target in can- serine or synthesize it de novo. Recently, overexpression of cers, but also provides a tool with which to study serine metabolism. 3-phosphoglycerate dehydrogenase (PHGDH), the gene encod- ing the first committed step of serine synthesis, via focal am- plification and other mechanisms, has been identified in human PHGDH | inhibitor | serine | cancer metabolism cancers. Cancer cell lines that overexpress PHGDH are uniquely sensitive to PHGDH knockdown whereas lines that express little Serine is required for a plethora of anabolic processes. Serine PHGDH are insensitive, suggesting that PHGDH may be a clini- is an abundant component of proteins and is required for the cally interesting target. Here, we report the discovery of a synthesis of , including sphingolipids and phosphatidylserine, specific small molecule inhibitor of PHGDH, which enables pre- a major component of cellular membranes (1–3). Alternatively, clinical evaluation of PHGDH as a target in cancer and provides a serine hydroxymethyltransferases (SHMTs) convert serine to gly- tool to study the biology of de novo serine synthesis. cine, concomitantly charging the folate pool with “one-carbon” units (4, 5). Both glycine and folate one-carbon units are used to Author contributions: E.M., N.C.L., L.L.L., and L.C.C. designed research; E.M., N.C.L., R.B.Z., make nucleotides. Thus, serine serves numerous critically impor- J.L.A., A.P.G., B.N.N., J.C.Y.W., S.C., H.T., P.K.S., J.D.W., S.-M.F., J.M.A., G.M.D., and C.A.L. tant roles in cellular metabolism. performed research; E.M., N.C.L., R.B.Z., J.L.A., A.P.G., B.N.N., S.C., H.T., P.K.S., J.B., J.D.W., At the cellular level, serine can be imported from the extra- S.-M.F., J.M.A., G.M.D., C.A.L., L.L.L., and L.C.C. analyzed data; and E.M., C.A.L., L.L.L., cellular space via amino acid transporters (6, 7). Alternatively, and L.C.C. wrote the paper. serine can be synthesized from glucose via the phosphoserine Reviewers: E.G., Beatson Institute for Cancer Research; and B.R.S., Columbia/Howard Hughes Medical Institute. pathway (8). De novo synthesis proceeds from the glycolytic intermediate 3-phosphoglycerate (3-PG) via three sequential en- Conflict of interest statement: L.C.C. owns equity in, receives compensation from, and serves on the Board of Directors and Scientific Advisory Board of Agios Pharmaceuticals. zymatic reactions (Fig. 1A), the first of which is catalyzed by + Agios Pharmaceuticals is identifying metabolic pathways of cancer cells and developing the NAD -dependent enzyme 3-phosphoglycerate dehydrogenase drugs to inhibit such to disrupt tumor cell growth and survival. (PHGDH) (9). For decades, it has been known that cancer cells 1Present address: Frontier Research Laboratories, Daiichi Sankyo Co., Shinagawa-ku, Tokyo have enhanced serine synthesis, which contributes to nucleotide 140-8710, Japan. synthesis (10, 11). Recently, focal amplifications of the gene encoding 2To whom correspondence may be addressed. Email: [email protected], clyssiot@ PHGDH have been reported, particularly in breast cancers and med.umich.edu, or [email protected]. melanomas (12–14). Additionally, KEAP1 and nuclear factor This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. erythroid-2-related factor 2 (NRF2) mutant non-small cell lung 1073/pnas.1521548113/-/DCSupplemental. www.pnas.org/cgi/doi/10.1073/pnas.1521548113 PNAS Early Edition | 1of6 Glucose Fig. 1. A B Screening Assay Screening for inhibitors of PHGDH. (A)Serine The Phosphoserine Pathway synthesis from glucose via the phosphoserine pathway: PHGDH PSAT1 Phosphoglycerate dehydrogenase (PHGDH) oxidizes 3PG p-Pyr p-Ser the glycolytic intermediate 3-phosphoglycerate (3-PG) to PHGDH PSAT1 PSPH + NADH Glu αKG 3PG P-Pyr P-Ser Ser NAD 3-phosphohydroxypyruvate (p-Pyr) using NAD+; phos- NAD+ NADH Glu αKG Diaphorase Resazurin phoserine amino (PSAT1) transaminates p-Pyr to phosphoserine (p-Ser) using glutamate as a ni- (Ex550/Em585) trogen donor; phosphoserine phosphatase (PSPH) de- Pyruvate Primary Screen phosphorylates p-Ser to yield serine. (B) In vitro PHGDH 800K compounds C D assay. Diaphorase couples the NADH produced upon PHGDH turnover to the reduction of resazurin to fluo- 3,906 Hits rescent resorufin. Resorufin fluorescence is a proxy for (Z score < -3) PHGDH activity. PSAT1 is included to prevent product feedback inhibition of PHGDH by p-Pyr. (C) Z-score plot Z-score Diaphorase for the 800,000-compound library screened using the (triplicate) counter screen above PHGDH assay. Each point represents a single compound. A negative score indicates inhibition. (D) Screen triaging strategy. Setting a Z-score threshold − Compounds of 3 gave 3,906 putative hits. After counter-screening Hits: 408 against diaphorase to rule out false positives and con- E CBR-5884 CBR-5807 CBR-6936 CBR-9480 firming activity against PHGDH, 408 compounds re- O O H S mained. Selected compounds were profiled against a N S S S N + O S N O N S O H panel of metabolic NAD(P) dehydrogenases to ascer- O O N SH (dehydrogenase panel) S S H HN tain selectivity for PHGDH. (E) Structures of represen- N tative PHGDH inhibitors evaluated in cell-based assays. inhibitors as a targeted therapy for these tumor types represent an (Table S1). A number of these compounds are likely to target exciting clinical opportunity. sulfhydryl groups and may therefore react with a PHGDH cys- The studies herein detail our efforts in identifying small mole- teine residue. For example, both CBR-5807 and CBR-6936 con- cule inhibitors of PHGDH. We reasoned that a PHGDH inhibitor tain sulfhydryl-reactive centers. Interestingly, CBR-5807 would have the benefits of not only providing a tool compound (Disulfiram) is an approved drug dosed in humans to treat alco- with which to study the biology of serine synthesis, but also en- holism and known to inhibit aldehyde dehydrogenase by reacting abling thorough preclinical evaluation of PHGDH as a target in with sulfhydryl groups (18). cancers. We screened a library of 800,000 small molecules using an in vitro PHGDH assay. A cell-based assay for serine synthesis was CBR-5884 Inhibits Serine Synthesis in Cells. We determined whether used to identify a lead, CBR-5884, that was active in cells. CBR- any of our seven leads inhibited serine synthesis in cancer cells. 5884 selectively inhibited the proliferation of melanoma and breast To do so, we turned to gas chromatography mass-spectrometry 13 cancer lines that have a high propensity for serine synthesis but had (GCMS) with uniformly carbon-13–labeled glucose ( C6-glucose) no effect on lines that rely on extracellular serine uptake. Mech- tracing. Given the isotopic enrichment of serine, it is possible to anistically, CBR-5884 was found to be a noncompetitive inhibitor, decouple newly synthesized serine from extracellular serine or showed a time-dependent onset of inhibition, and disrupted the serine that was synthesized before tracer addition. Newly syn- oligomerization state of PHGDH. thesized serine has a mass-shift of 3 (M+3) due to the incorpora- tion of glucose-derived 13C via 3-PG. We first investigated the Results kinetics of serine labeling. Serine labeling plateaued around 6 h, Screening for Small Molecule Inhibitors of PHGDH. An in vitro en- with ∼65% of the serine pool being 13C-labeled (Fig. S1). The zymatic assay for PHGDH activity amenable to high-throughput plateau phase likely reflects exchange between intra- and extra- screening (HTS) was developed by coupling the production of cellular serine pools (19). NADH, upon 3-PG oxidation, to the reduction of resazurin to With an understanding of serine-labeling kinetics, we designed B 13 resorufin using diaphorase as the coupling enzyme (Fig. 1 ). Thus, a C6-glucose tracing assay to acutely interrogate the effects of resorufin fluorescence served as a proxy for PHGDH activity. compounds on serine synthesis (Fig. 2A). Assaying serine syn- The assay was miniaturized to a 1,536-well format with a thesis with a 3-h compound treatment was preferred to longer Z-factor of >0.75, indicating a high quality assay (17). A library treatments to guard against false positives that decrease serine of 800,000 small molecules was screened in single point format labeling by an indirect effect, such as generally compromising at 13 μM(Fig.1C). Setting a threshold Z score of −3, corre- cellular viability. Among our lead compounds, CBR-5884 was able sponding to at least 50% PHGDH inhibition, gave a 0.5% hit rate, to decrease de novo serine synthesis by 30%; the remaining com- yielding 3,906 hits. Putative hits were reassayed in triplicate and pounds had little effect (Fig. 2B). The dose at which CBR-5884 had counter-screened against diaphorase to rule out false positives an effect on serine labeling was consistent with the in vitro biochemical targeting diaphorase. The counter screen eliminated 3,498 com- IC of 33 ± 12 μM for PHGDH (Fig. 2C). At such concentrations, 50 + pounds, giving 408 PHGDH inhibitors (Fig. 1D). CBR-5884 had no effect on two other NAD -dependent de- A triaging strategy based on hit potency and selectivity was hydrogenases, (LDH) and MDH1 (Fig. designed. We reasoned that inhibitors specific to PHGDH would 2C and Fig. S1). Importantly, under the acute treatment time pe- minimize general cellular toxicity compared with compounds that riod used in the labeling assays, CBR-5884 was not generally cytotoxic hit a variety of dehydrogenases. Thus, half maximal inhibitory at concentrations up to 40 μM as determined by two independent + concentrations (IC50) were determined for a panel of NAD(P) - cellular viability assays (Fig. S1). Therefore, decreases in serine la- dependent dehydrogenases that included PHGDH, isocitrate beling are a direct effect of CBR-5884–mediated PHGDH inhibition. dehydrogenase (IDH1), (MDH1), and 3α- We resynthesized CBR-5884 in-house and performed a dose– hydroxysteroid dehydrogenase (3α-HSD). Compounds at least response experiment for CBR-5884 using the same acute treat- fourfold more selective for PHGDH were progressed for fur- ment method as above. Serine labeling was significantly decreased ther analysis. Based on this triaging, seven of the most potent at 30 μM and trended toward a decrease at 15 μM(Fig.2D). Im- PHGDH inhibitors were selected as lead compounds for evalua- portantly, perturbations in labeling were specific to serine in that tion in cell-based assays; selected structures are shown in Fig. 1E neither the PHGDH substrate, 3-PG, nor the end products of

2of6 | www.pnas.org/cgi/doi/10.1073/pnas.1521548113 Mullarky et al. glycolysis, pyruvate and lactate, were affected (Fig. 2D). We further PHGDH expression (Fig. S2); MDA-MB-468 and HCC70 cells confirmed that glycolytic metabolites were unperturbed by CBR- harbor PHGDH amplifications (14). In contrast, serine depletion 5884 treatment using liquid chromatography mass spectrometry almost completely abrogated proliferation of low PHGDH- (LC-MS/MS) to interrogate a greater panel of metabolites (Fig. expressing lines MDA-MB-231 and MCF10A (Fig. 3B). In mel- S1). Thus, changes in serine labeling are a direct effect of CBR- anoma cells, PHGDH protein levels were similarly commen- 5884–mediated PHGDH inhibition and not a consequence of surate with the ability to proliferate in serine-free media (Fig. changes in PHGDH substrate levels or general perturbations in S2). Interestingly, although Carney cells are sensitive to extra- glycolytic flux (Fig. 2D). The absence of an effect on lactate la- cellular serine depletion, they can adapt and proliferate, as evi- beling was consistent with the in vitro data showing that CBR-5884 denced by increased PHGDH protein levels upon serine depletion does not inhibit LDH under the drug concentrations used. In sum, (Fig. S2). the data argue that CBR-5884 is able to selectively inhibit serine Given that the ability to proliferate in the absence of extra- synthesis in cells. cellular serine is indicative of a high propensity for serine syn- Given that CBR-5884 is an ethyl ester and therefore suscep- thesis, we hypothesized that such lines should be sensitive to tible to intracellular esterases, we investigated whether the car- CBR-5884. Conversely, lines that cannot grow in serine-free media boxylic acid derivative of the parent molecule was still active have a low propensity for serine synthesis and should therefore be against PHGDH; were the acid less active, it would likely de- resistant to PHGDH inhibition. Treating the breast lines with CBR- crease the efficiency of targeting PHGDH in situ. Parent and 5884 in serine-replete media inhibited growth of the four lines that acid derivatives were equally potent against, and selective for, grew without extracellular serine in a dose-dependent manner, with PHGDH in vitro, suggesting that intracellular deesterification is growth inhibition ranging from 35% to 60% at 30 μM CBR-5884. unlikely to affect CBR-5884 activity (Fig. S1). The inhibitor had no effect on the three lines sensitive to serine withdrawal, indicating that the inhibitor was selectively toxic to cells CBR-5884 Selectively Inhibits the Proliferation of Cancer Cells with a with high serine synthesis activity (Fig. 3C). We next asked whether High Propensity for Serine Synthesis. We established a system to removing serine from the media, to enhance the reliance on de test the ability of CBR-5884 to inhibit PHGDH-dependent cancer novo serine synthesis, could sensitize cells to PHGDH inhibition. cell proliferation. We first evaluated the ability of a panel of breast Indeed, serine depletion increased the efficacy of CBR-5884 in lines and melanoma cell lines to proliferate in serine replete or deplete already sensitive under serine-replete conditions as evidenced by an media as a proxy for serine biosynthetic activity. Breast lines were 80–90% decrease in proliferation with 30 μMCBR-5884(Fig.3D). selected based on PHGDH expression according to the Cancer Cell Moreover, MCF7 cells, which were of intermediate sensitivity to

Line Encyclopedia (CCLE) data and validated by blotting for serine withdrawal, and insensitive to drug under serine-replete BIOCHEMISTRY PHGDH (Fig. 3A and Fig. S2) (20). Removing extracellular serine conditions, became partially sensitive to the inhibitor under serine- hadnoeffectonproliferationofhighPHGDH-expressing lines deplete conditions (Fig. 3D). Importantly, under serine-replete con- MDA-MB-468, MDA-MB-436, HCC70, and Hs578T (Fig. 3B). All ditions, PHGDH knockdown phenocopied the effects of CBR-5884 four lines cluster in the top quartile of the CCLE dataset for treatment in that the drug-sensitive lines were also sensitive to

A Pre-treatment Labeling t = -1h t = 0 t = 2h

Add drug Wash, Harvest 13 add C6glucose metabolites and drug BC 0.6 1.00

* 0.75 0.4 o PHGDH / V M+3 ) LDH i V 0.50 0.2 (Fraction IC50 = 33 ± 12 M Glucose-derived serine 0.25 0.0 Fig. 2. CBR-5884 inhibits serine synthesis in cells. 7 6 4 9 -3 -6 -0 (A) Acute inhibitor treatment assay schematic. Carney 580- 740- 588- 743- 690 693 948 0.00 DMSO -7 -6 -5 -4 -3 cells are pretreated with drug at 30 μMfor1hbefore 13 Inhibitor LOG [CBR-5884] initiating C6-glucose labeling for 2 h still in the D presence of drug. Polar metabolites are harvested 1.0 3PG 0.6 Serine 1.0 Pyruvate 1.0 Lactate andanalyzedbyGCMS.(B) The ability of the seven lead PHGDH inhibitors to block serine synthesis was 13 assayed as in A. C6-glucose–derived serine (M+3 0.8 0.8 0.8 serine) relative to total serine levels is plotted. (C)In * 0.4 vitro IC50 assays for PHGDH and lactate dehydrogenase 0.6 0.6 0.6 (LDH). Initial rates of the enzymatic reaction (Vi)atthe indicated CBR-5884 concentration normalized to that

of the DMSO control (Vo) are plotted. (D) CBR-5884 tion M+3 ) 0.4 0.4 0.4 0.2 dose–response experiment as in A, but monitoring a

(Frac panel of phosphoserine pathway and glycolytic metab- 0.2 0.2 0.2 olites. The y axes indicate the fraction of the indicated Glucose-derived metabolite metabolite derived from glucose (M+3 metabolite 0.0 0.0 0.0 0.0 level relative to total metabolite level). Asterisks in- M M M M M M M M M M M M M M M M < μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ dicate significant differences vs. DMSO treatment (P 0 1 0 1 0 1 0 1 15 30 15 30 15 30 15 30 0.05, t test, n ≥ 3). Error is given as ±1SD.

Mullarky et al. PNAS Early Edition | 3of6 PHGDH knockdown (Fig. 3E and Fig. S2). Furthermore, as which is resistant to CBR-5884–mediated inhibition (Fig. S3). with the drug treatments, growing cells in serine-free media CBR-5884 still inhibited a truncated form of PHGDH, which enhanced the growth inhibitory effect of PHGDH knockdown lacks the C-terminal domain responsible for tetramerization and (Fig. 3F). Similar trends were observed for the melanoma panel is therefore a constitutive dimer (Fig. S3). Together, these results in terms of both the selectivity of CBR-5884 for cells with a high suggest that disruption of the tetramer might assist PHGDH in- propensity for serine synthesis and the increased efficacy under hibition but is not necessary for inhibition. serine-deplete conditions (Fig. S2). Finally, the acid derivative of compound 5884 was not effective on MDA-MB-468 cells, Discussion likely owing to poor membrane permeability, and is therefore We have reported the discovery of a PHGDH inhibitor, CBR- not a viable alternative to the parent compound (Fig. S2). 5884, and have shown that it inhibits serine synthesis in cells. Furthermore, CBR-5884 specifically inhibited the proliferation of Analysis of CBR-5884 Inhibition Modality. We sought to more deeply melanoma and breast cancer lines with high levels of serine syn- characterize the mechanism by which CBR-5884 inhibits PHGDH. thesis activity, with little effect on lines reliant on serine import. Inhibition constants (Ki) for CBR-5884 with respect to each Thus, CBR-5884 is selective for lines addicted to serine synthesis substrate were determined. CBR-5884 inhibited PHGDH in and phenocopies sensitivity to PHGDH knockdown. Finally, a a noncompetitive mode with respect to both substrates, as evi- biochemical analysis of CBR-5884 revealed that it was a non- denced by a decreasing V with increasing CBR-5884 con- max competitive inhibitor that showed a time-dependent onset of in- centration. The inhibition constants were 50 ± 20 μM and 50 ± + 3 μM for 3-PG and NAD , respectively (Fig. 4 A and B). We hibition and disrupted the oligomerization state of PHGDH. assessed whether there was any time dependence to the onset of Recent work examining how malignant cells rewire their me- inhibition by varying the time period for which drug and tabolism to support growth and proliferation has revealed a number PHGDH were preincubated before initiating the enzymatic re- of clinically interesting targets (21, 22). Perhaps the most promising action. CBR-5884 was progressively more potent with increasing is the discovery of gain-of-function mutations in the (IDH) enzymes that result in the production of preincubation time, culminating in an IC50 of 7 μM when drug and PHGDH were preincubated for 4 h (Fig. 4C). Intrigued the oncometabolite 2-hydroxyglutarate (23, 24). The findings trans- by the combination of a time-dependent onset of inhibition and lated into chemical probes that yielded insights into the biology noncompetitive inhibition, the latter suggesting that CBR-5884 of IDH mutations (25, 26) and led to clinical programs (e.g., might be binding to an allosteric pocket, we speculated that CBR- NCT02481154). The genetic evidence pointing to a role for PHGDH 5884 could be affecting the PHGDH oligomerization state, where in cancer is similarly striking: PHGDH is one of few metabolic the time dependency of inhibition could potentially stem from enzymes genetically deregulated in cancer (27, 28). Notably, el- drug-induced conformational changes in PHGDH. To evaluate evated PHGDH expression correlates with clinical aggressiveness the PHGDH oligomerization state, we incubated PHGDH and poor prognosis in TNBC (13, 16) and NSCLC (15). There is a with drug and then cross-linked before SDS/PAGE. CBR-5884 paucity of targeted therapies for these cancers and chemother- shifted the PHGDH equilibrium from the tetrameric to the apies are frequently used (29, 30). Thus, the clinical potential of dimeric state (Fig. 4D). No such effect was observed with LDH, PHGDH inhibitors as targeted agents for TNBC and NSCLC

A B +SER -SER

A-MB-231 1.00 MDA-MB-468MDA-MB-436Hcc70 Hs578TMCF-7 MD MCF-10A 0.75 PHGDH (light) 0.50 PHGDH (dark) 0.25

Vinculin number cell Relative 0.00

Hcc70 Hs578T MCF-7 MCF-10A MDA-MB-468MDA-MB-436 MDA-MB-231

CDDMSO 15μM 30μM DMSO 15μM 30μM

1.00 1.00

0.75 0.75

0.50 0.50 Fig. 3. 0.25 0.25 CBR-5884 selectively inhibits the proliferation Relative cell number cell Relative Relative cell number cell Relative of breast cancer lines with a high propensity for ser- 0.00 0.00 ine synthesis. (A) Western blot for lines grown in +SER

cc70 -231 -468 cc70 H MCF-7 B-436 H MCF-7 media: two lanes per cell line with each lane loaded Hs578T -MB -M Hs578T MCF-10A A DA MDA-MB-468 MDA-MB-436 M MDA-MB MD with independent cell lysates. (B) Proliferation assay for breast lines grown in either serine-replete (+SER) EFor -deplete (−SER) media. Proliferation assay for lines shGFP sh1 sh2 shGFP sh1 sh2 1.25 treated with CBR-5884 in (C) +SER or (D) −SER 1.00 + 1.00 media. Proliferation assay for lines grown in (E) SER 0.75 or (F) −SER media with PHGDH knockdown (sh1 and 0.75 sh2) or a nontargeting control (shGFP). MDA-MB-468 0.50 0.50 and HCC70 are PHGDH amplified. MCF-10A cells are nontransformed mammary epithelial cells; other lines Relative cell number cell Relative 0.25 0.25

Relative cell number cell Relative are cancer cell lines. MDA-MB-231 and MCF-10A lines 0.00 0.00 were not included in −SER experiments in D and F -436 MCF-7 B-231 B MCF-7 -MB-436 Hs578T -M Hs578T because they are sensitive to serine withdrawal. His- A MCF-10A DA DA-M MDA-MB-468 M MD MDA-MB-468 M tograms depict mean ± SE (n ≥ 3).

4of6 | www.pnas.org/cgi/doi/10.1073/pnas.1521548113 Mullarky et al. tumors addicted to serine synthesis, as a single agent or in com- provide a proof-of-concept that small molecule inhibitors of bination with standard of care, is an exciting perspective. PHGDH represent a viable class of anti-cancer drugs. Beyond the preclinical applications, a PHGDH inhibitor provides a tool to study de novo serine synthesis. For example, it remains Methods unclear why a serine biosynthesis enzyme is critical for tumor growth Cells, Transfections, and Infections. Breast and melanoma lines were passaged in when serine is available in the serum (13, 14). CBR-5884 provides a RPMI supplemented with 10% (vol/vol) FBS, penicillin, streptomycin, and nor- valuable tool complementary to genetic strategies to study the ne- mocin (InvivoGen). Lentivirus was produced from Lenti-X 293T cells (Clontech) cessity of serine synthesis and other phenomena because small transfected with packaging plasmids pCMV-dR8.2 and pCMV-VSV-G and in- molecules provide greater temporal resolution and do not deplete dicated pLKO.1 shRNAs: PHGDH, TRCN0000233029 (sh1) and TRCN0000221864 the actual protein. Moreover, CBR-5884 has already suggested an (sh2); nontargeting control, TRCN0000072181 (shGFP). Experiments were per- interesting feature of PHGDH biochemistry: namely, that human formed in compliance with Weill Cornell Medicine Environmental Health and PHGDH could be regulated by transitions between different olig- Safety and Institutional Biosafety Committee. See SI Methods for details. omerization states. These findings are reminiscent of pyruvate Immunoblots. kinase M2 (PKM2) regulation in that both an endogenous me- Protein was extracted from cells via trichloroacetic acid pre- cipitation and blotted for with primary antibodies αPHGDH (HPA021241, 1/10,000; tabolite, fructose-1,6-bisphosphate, and pharmacological small Sigma) and αVinculin (V9264, 1/5,000; Sigma). See SI Methods for details. molecule activators enhance PKM2 activity by stabilizing the tetrameric form (31, 32). Endogenously, 2-phosphoglycerate (2- Proliferation Assays. Cells were plated at a low density in 96- or 24-well plates PG) has been reported to activate PHGDH (33). Mechanistically, in serine containing media. The following day, media were aspirated, cells 2-PG could be functioning by modulating the PHGDH oligo- were washed with PBS, and fresh serine-replete or -deplete media containing merization state. Finally, it is possible that CBR-5884 is a co- drug (15 μM, 30 μM) or vehicle (DMSO) were added. Cells were grown for valent inhibitor of PHGDH. Although PHGDH does not rely on 3–5 d, with drug and media changed daily before assaying relative cell an cysteine, there are a number of cysteines that could numbers. See SI Methods for details. potentially perturb enzymatic function were they modified because 13 they are in close proximity to the active site (PDB ID code 2G76). Acute Drug Treatments with C6-Glucose Tracing. Carney cells acclimated to In an attempt to determine whether changes in the PHGDH growth in MEM (Corning) were plated at 9 × 105 cells per 6-cm dish the night oligomerization state are unique to CBR-5884, it would be in- before. The following morning, media were replaced with fresh media teresting to determine whether Disulfiram (CBR-5807), CBR- containing CBR-5884 (1 μM, 15 μM, 30 μM) or vehicle control (DMSO) for 1 h. 6936, or sulfhydryl blocking reagents have similar effects. Media were then aspirated, cells were washed with PBS, and fresh glucose- 13 The identification of a selective small molecule inhibitor of free MEM (Gibco) supplemented with C6-glucose (3 g/L; Cambridge Iso- BIOCHEMISTRY PHGDH capable of modulating de novo serine synthesis in topes) and 10% (vol/vol) dialyzed FBS containing drug or DMSO was added. PHGDH-dependent cancer cells represents a significant step to- After 2 h, cells were quickly washed with cold PBS on ice and flash frozen. ward the goal of targeting serine metabolism in oncology. How- Polar metabolites were extracted as in the GCMS methods. See SI Methods ever, future in vivo evaluation of the CBR-5884 chemical series for details. will require medicinal chemistry-based optimization. Indeed, CBR- Acute Toxicity Assay. Carney cells acclimated to growth in MEM media were 5884 was found to be unstable in mouse plasma, and, as described, plated in a 96-well plate at 6,000 cells per well. The next day, cells were treated replacement of the ethyl ester moiety with the corresponding with CBR-5884 from 1 μMto40μM for 3 h. Drug containing media were then negatively charged carboxylic acid resulted in a derivative that removed, fresh drug-free media added, and cell viability was determined via a retains enzyme inhibitory activity but loses activity on cells. Thus, CellTiter-Glo (G7572; Promega) or Alamar Blue (DAL1025; Invitrogen) assay CBR-5884 is more likely to serve as a tool compound, and as a according to the manufacturer’s protocol. See SI Methods for details. starting point for generating more drug-like molecules, than an actual drug. Furthermore, the cell-based potency of this series will GCMS Metabolite Analysis. Polar metabolites were extracted with 2 mL of likely need to be improved to enable in vivo evaluation at expo- MeOH/H2O (4:1) for 30 min on dry ice, scraped, transferred to 2-mL tubes, sure levels that are not generally toxic. Nevertheless, our results and centrifuged (30 min, 21,000 × g), and the supernatants were dried under

AB

Fig. 4. Mechanisms of CBR-5884 inhibition. Inhibition

constants (Ki) were determined by titrating (A)3-PGor (B) NAD+ while holding the other substrate constant at four different CBR-5884 concentrations and de- termining the initial reaction rate using a PHGDH assay. CD Plots were fit to a noncompetitive model. (C)Time- dependent inhibition was measured by preincubating drug and PHGDH for 0.5 h, 1 h, or 4 h as indicated before initiating the PHGDH reaction. Initial reaction

rates (Vi) were determined and normalized to that of DMSO (Vo). (D) PHGDH was preincubated with CBR- 5884 (0, 50, 200, or 400 μM) before cross-linking with BS3 (0.25 or 2.5 mM) followed by SDS/PAGE and Coomassie staining. Leftmost lane had no BS3 indi- cating the monomeric species. Oligomerization state was inferred from reference to a molecular weight ladder. Error is given as ±1SD(n ≥ 3).

Mullarky et al. PNAS Early Edition | 5of6 vacuum. Samples were derivatized as previously described (34) and analyzed Cross-Linking Assays. PHGDH (1.5 μg) or LDH (2.2 μg, 59747; Sigma) was in- on an Agilent 6890 GC instrument. Metabolite quantification was inferred cubated with CBR-5884 (50 μM, 200 μM, 400 μM) or vehicle control (DMSO) from a standard curve, and fractional enrichment of 13C in metabolites was in 25 mM Hepes, pH 7.3, and 1 mM NAD+ in 18 μL total volume for 30 min corrected for the natural abundance of 13C and 15N (35, 36). See SI Methods before BS3 (Pierce) cross-linking and quenching. Samples were run on SDS/ for details. PAGE and colloidal Coomassie stained (Bio-Rad). See SI Methods for details.

LC-MS/MS Metabolite Analysis. Polar metabolites were extracted and dried as Primary PHGDH Screen, Diaphorase Counter Screen, and Dehydrogenase Panel in the GCMS method. Samples were resuspended in 15 μL of HPLC grade Selectivity Profiling. Compounds (800,000) were screened at a single dose water. Then, 5 μL of each sample was injected and analyzed using a 5500 (13.3 μM) in 1,536-plate format against PHGDH or diaphorase quantifying QTRAP triple quadrupole mass spectrometer (AB/Sciex) coupled to a Prom- resorufin fluorescence (550/590 nm; Ex/Em) with an Envision plate reader. inence UFLC system (Shimadzu) as reported previously (37). Results were analyzed using Genedata Screener software. Compounds with a robust Z-score of <−3 in the PHGDH screening assay and robust Z-score of >− Protein Purification. His6-tagged pET28a PHGDH, pET28a PSAT1, and pNIC28- 2 in the diaphorase counter screen were selected as hits. See SI Methods Bsa4 PHGDH3-314 were purified via nickel agarose (Qiagen) from BL21 for detailed protocols, hit selection and confirmation, and selectivity pro- Escherichia coli cultures. pVB-CBD IDH1 was purified via Macroporous Bead filing against the dehydrogenase panel. Cellulose capture, TEV protease (Sigma-Aldrich) digestion, and gel filtration chromatography from BL21 E. coli cultures. See SI Methods for details. Chemical Syntheses. CBR-5884, ethyl 5-(furan-2-carboxamido)-3-methyl-4- thiocyanatothiophene-2-carboxylate, and the acid derivative, 5-(furan-2- PHGDH Assays. PHGDH activity was measured in 96-well plates by monitoring carboxamido)-3-methyl-4-thiocyanatothiophene-2-carboxylic acid, synthesis was NADH fluorescence [excitation wavelength (Ex) 340 nm/emission wavelength adapted from the literature as described in SI Methods (38, 39). (Em) 460 nm] over time. PSAT1 was included to prevent product inhibition of PHGDH. See SI Methods for details. ACKNOWLEDGMENTS. We thank U. Oppermann, M. G. Vander Heiden, K. R. Mattaini, and M. Yuan for technical assistance and reagents. We thank LDH and MDH1 Assays. Enzyme activities were assayed using kits (for LDH, J. Johnson, Y. Zheng, H. Shim, B. D. Ngo, and other L.C.C. laboratory members for helpful discussions. L.C.C. was supported by NIH Grants P01CA117969 and MAK06, Sigma; for MDH1, MAK196-1KT, Sigma) according to the manufac- ’ P01CA120964. C.A.L. was partially supported by a PanCAN-AACR Pathway to turer s instructions with commercially available recombinant enzyme (for LDH, Leadership Award and a Dale F. Frey Award for Breakthrough Scientists from 59747, Sigma; for MDH1, SRP6103, Sigma). Drug, titrated as for the PHGDH the Damon Runyon Cancer Research Foundation (Grant DFS-09-14). G.M.D. IC50 assays, and enzyme were preincubated for 30 min before initiating re- was supported by a PanCAN-AACR Pathway to Leadership Award. S.-M.F. was action with substrate. supported by a Conquer Cancer Now Award from the Concern Foundation.

1. Kuge O, Hasegawa K, Saito K, Nishijima M (1998) Control of phosphatidylserine 20. Barretina J, et al. (2012) The Cancer Cell Line Encyclopedia enables predictive mod- biosynthesis through phosphatidylserine-mediated inhibition of phosphatidylserine elling of anticancer drug sensitivity. Nature 483(7391):603–607. synthase I in Chinese hamster ovary cells. Proc Natl Acad Sci USA 95(8):4199–4203. 21. Son J, et al. (2013) Glutamine supports pancreatic cancer growth through a KRAS- 2. de Koning TJ, et al. (2003) L-serine in disease and development. Biochem J 371(Pt 3): regulated metabolic pathway. Nature 496(7443):101–105. 653–661. 22. Tennant DA, Durán RV, Gottlieb E (2010) Targeting metabolic transformation for 3. Futerman AH, Riezman H (2005) The ins and outs of sphingolipid synthesis. Trends Cell cancer therapy. Nat Rev Cancer 10(4):267–277. Biol 15(6):312–318. 23. Dang L, et al. (2009) Cancer-associated IDH1 mutations produce 2-hydroxyglutarate. 4. Stover P, Schirch V (1990) Serine hydroxymethyltransferase catalyzes the hydrolysis of Nature 462(7274):739–744. 5,10-methenyltetrahydrofolate to 5-formyltetrahydrofolate. J Biol Chem 265(24): 24. Parsons DW, et al. (2008) An integrated genomic analysis of human – 14227 14233. multiforme. Science 321(5897):1807–1812. 5. Tibbetts AS, Appling DR (2010) Compartmentalization of mammalian folate-mediated 25. Wang F, et al. (2013) Targeted inhibition of mutant IDH2 in leukemia cells induces – one-carbon metabolism. Annu Rev Nutr 30(1):57 81. cellular differentiation. Science 340(6132):622–626. 6. Palacín M, Estévez R, Bertran J, Zorzano A (1998) Molecular biology of mammalian 26. Rohle D, et al. (2013) An inhibitor of mutant IDH1 delays growth and promotes dif- – plasma membrane amino acid transporters. Physiol Rev 78(4):969 1054. ferentiation of cells. Science 340(6132):626–630. 7. Barker GA, Ellory JC (1990) The identification of neutral amino acid transport systems. 27. Mullen AR, DeBerardinis RJ (2012) Genetically-defined metabolic reprogramming in – Exp Physiol 75(1):3 26. cancer. Trends Endocrinol Metab 23(11):552–559. 8. Snell K (1986) The duality of pathways for serine biosynthesis is a fallacy. Trends 28. Gottlieb E, Tomlinson IPM (2005) Mitochondrial tumour suppressors: A genetic and Biochem Sci 11(6):241–243. biochemical update. Nat Rev Cancer 5(11):857–866. 9. Mullarky E, Mattaini KR, Vander Heiden MG, Cantley LC, Locasale JW (2011) PHGDH 29. Foulkes WD, Smith IE, Reis-Filho JS (2010) Triple-negative breast cancer. N Engl J Med amplification and altered glucose metabolism in human melanoma. Pigment Cell 363(20):1938–1948. Melanoma Res 24(6):1112–1115. 30. Boolell V, Alamgeer M, Watkins DN, Ganju V (2015) The evolution of therapies in non- 10. Snell K, Natsumeda Y, Eble JN, Glover JL, Weber G (1988) Enzymic imbalance in serine small cell lung cancer. Cancers (Basel) 7(3):1815–1846. metabolism in human colon carcinoma and rat sarcoma. Br J Cancer 57(1):87–90. 31. Anastasiou D, et al. (2012) Pyruvate kinase M2 activators promote tetramer formation 11. Snell K, Natsumeda Y, Weber G (1987) The modulation of serine metabolism in and suppress tumorigenesis. Nat Chem Biol 8(10):839–847. hepatoma 3924A during different phases of cellular proliferation in culture. Biochem 32. Kung C, et al. (2012) Small molecule activation of PKM2 in cancer cells induces serine J 245(2):609–612. auxotrophy. Chem Biol 19(9):1187–1198. 12. Beroukhim R, et al. (2010) The landscape of somatic copy-number alteration across 33. Hitosugi T, et al. (2012) Phosphoglycerate mutase 1 coordinates glycolysis and bio- human cancers. Nature 463(7283):899–905. – 13. Locasale JW, et al. (2011) Phosphoglycerate dehydrogenase diverts glycolytic flux and synthesis to promote tumor growth. Cancer Cell 22(5):585 600. contributes to oncogenesis. Nat Genet 43(9):869–874. 34. Nicolay BN, et al. (2013) Loss of RBF1 changes glutamine catabolism. Dev 27(2): – 14. Possemato R, et al. (2011) Functional genomics reveal that the serine synthesis 182 196. pathway is essential in breast cancer. Nature 476(7360):346–350. 35. Antoniewicz MR, Kelleher JK, Stephanopoulos G (2006) Determination of confidence 15. DeNicola GM, et al. (2015) NRF2 regulates serine biosynthesis in non-small cell lung intervals of metabolic fluxes estimated from stable isotope measurements. Metab – cancer. Nat Genet 47(12):1475–1481. Eng 8(4):324 337. 16. Pollari S, et al. (2011) Enhanced serine production by bone metastatic breast cancer 36. Antoniewicz MR, Kelleher JK, Stephanopoulos G (2007) Elementary metabolite units cells stimulates osteoclastogenesis. Breast Cancer Res Treat 125(2):421–430. (EMU): A novel framework for modeling isotopic distributions. Metab Eng 9(1):68–86. 17. Zhang JH, Chung TD, Oldenburg KR (1999) A simple statistical parameter for use in 37. Ying H, et al. (2012) Oncogenic Kras maintains pancreatic tumors through regulation evaluation and validation of high throughput screening assays. J Biomol Screen 4(2): of anabolic glucose metabolism. Cell 149(3):656–670. 67–73. 38. Guzi JT, et al. (2007) Methods for inhibiting protein kinases. US Patent Appl 11/ 18. Vallari RC, Pietruszko R (1982) Human aldehyde dehydrogenase: Mechanism of in- 598,188:1–346. hibition of disulfiram. Science 216(4546):637–639. 39. Ambrogi V, Grandolini G, Perioli L, Rossi C (1992) Convenient synthesis of 2-amino- 19. Buescher JM, et al. (2015) A roadmap for interpreting (13)C metabolite labeling naphthalene-1-thiol and 3-aminoquinoline-4-thiol and cyclocondensations to 1,4- patterns from cells. Curr Opin Biotechnol 34:189–201. thiazino and 1,4-thiazepino derivatives. Synthesis 1992(7):656–658.

6of6 | www.pnas.org/cgi/doi/10.1073/pnas.1521548113 Mullarky et al.