sign in sign up
<<
Home , Amino acid, MTHFD1

Human SHMT inhibitors reveal defective import as a targetable metabolic vulnerability of diffuse large B- lymphoma

Gregory S. Duckera,b, Jonathan M. Ghergurovichb,c, Nello Mainolfid,VipinSurid, Stephanie K. Jeonga, Sophia Hsin-Jung Lic, Adam Friedmand, Mark G. Manfredid, Zemer Gitaic, Hahn Kima,e, and Joshua D. Rabinowitza,b,1

aDepartment of Chemistry, Princeton University, Princeton NJ 08544; bLewis-Sigler Institute for Integrative , Princeton University, Princeton NJ 08544; cDepartment of Molecular , Princeton University, Princeton NJ 08544; dRaze Therapeutics, Cambridge, MA 02139; and ePrinceton University Screening Center, Princeton University, Princeton, NJ 08544

Edited by Chi V. Dang, University of Pennsylvania School of Medicine, Philadelphia, PA, and accepted by Editorial Board Member Tak W. Mak September 6, 2017 (received for review April 20, 2017)

The hydroxymethyltransferse (SHMT) converts serine One- is targeted therapeutically by multiple into glycine and a tetrahydrofolate-bound one-carbon unit. existing , including the common clinical agents pemetrexed, one-carbon units support purine and thymidine synthesis, and thus 5-fluorouracil, and methotrexate (16). One mechanism of action . Mammals have both cytosolic SHMT1 and mitochon- common to several of these agents is inhibition of thymidylate syn- drial SHMT2, with the mitochondrial strongly up-regulated thase, which utilizes 5,10-methylene–THF. While new chemical tools in cancer. Here we show genetically that dual SHMT1/2 knockout have recently been disclosed that block de novo serine synthesis (17– blocks HCT-116 colon cancer tumor xenograft formation. Building 19), no existing chemotherapies specifically target the production of from a pyrazolopyran scaffold that inhibits plant SHMT, we identify 1C units from serine, the primary source of 1C units in tumors. ∼ small-molecule dual inhibitors of human SHMT1/2 (biochemical IC50 To block the production of 1C units from serine, simultaneous 10 nM). Metabolomics and isotope tracer studies demonstrate inhibition of both the cytosolic SHMT1 and mitochondrial SHMT2 effective cellular target engagement. A cancer cell-line screen is necessary. Here we genetically validate that dual SHMT1/2 genetic revealed that B-cell lines are particularly sensitive to SHMT inhibi- knockout, in Ras-driven colon cancer cells, prevents xenograft for- tion. The one-carbon donor formate generally rescues cells from ’ mation. We present the development of a low nanomolar, stereo- SHMT inhibition, but paradoxically increases the inhibitor s cyto- specific small-molecule inhibitor of human SHMT1/2. Dual SHMT toxicity in diffuse large B-cell lymphoma (DLBCL). We show that inhibition blocks growth of many cell lines in a manner that is res- this effect is rooted in defective glycine uptake in DLBCL cell lines, cued by the soluble 1C donor formate. In diffuse large B-cell lym- rendering them uniquely dependent upon SHMT enzymatic activ- ity to meet glycine demand. Thus, defective glycine import is a phoma (DLBCL) cell lines, however, formate does not rescue cell targetable metabolic deficiency of DLBCL. growth but instead paradoxically enhances cancer cell death. We find that this unexpected outcome reflects a previously unappreciated SHMT | cancer metabolism | glycine | DLBCL | folate biochemical vulnerability of DLBCL: inability of these cells to

ancer growth and proliferation are supported by metabolic Significance Cchanges, including enhanced uptake, aerobic (the Warburg effect), and folate-dependent one-carbon (1C) me- of the folate cycle are among the most consistently tabolism (1, 2). The predominant source of 1C units in cancer cells overexpressed in cancer. Whereas multiple clinical is the amino serine (3). The enzyme serine hydroxymethyl- agents inhibit thymidylate synthase, no current drugs target the (SHMT) catalyzes the conversion of serine and tetra- incorporation of one-carbon into via serine hydroxy- hydrofolate (THF) into glycine and 5,10-methylene–THF. Increases methyltransferase (SHMT). Using , we show that cancer in the synthesis and consumption of serine and glycine have been cells require SHMT to generate tumors.Wethendescribesmall- identified in transformed cells and cancers (4–6). Mitochondrial molecule SHMT inhibitors, and show that they block the growth SHMT (SHMT2) and the immediately downstream mitochon- of many human cancer cells, with B-cell lymphomas particularly drial enzyme 5,10-methylene-tetrahydrofolate dehydrogenase sensitive to SHMT inhibition. We find that this sensitivity arises (MTHFD2) are the most consistently overexpressed metabolic from the lymphomas’ inability to import the glycine, enzymesincancer(7–9) (Fig. 1A). In most rapidly proliferating which is made as a byproduct of the SHMT reaction. Thus, B-cell cells, 1C units generated from serine in the mitochondria lymphomas have an intrinsic defectinaminoacidimport,which areexportedtothecytosolasformate, which is then reassimilated causes a therapeutically targetable metabolic vulnerability. into folates to support synthesis (10–12). While the mitochondrial pathway typically supplies all of the 1C Author contributions: G.S.D., N.M., V.S., S.H.-J.L., A.F., M.G.M., Z.G., H.K., and J.D.R. de- signed research; G.S.D., J.M.G., S.K.J., and S.H.-J.L. performed research; N.M. and H.K. units in proliferating cells in culture, it is not essential in contributed new reagents/analytic tools; G.S.D. J.M.G., S.H.-J.L., and V.S. analyzed data; replete conditions, as evidenced by the viability of SHMT2 and and G.S.D. and J.D.R. wrote the paper. MTHFD2 deletion cell lines (11, 13). In such deletion cells, cytosolic Conflict of interest statement: N.M., V.S., A.F., and M.G.M. are employees of Raze Ther- SHMT1 now metabolizes serine to produce 1C units required for apeutics. J.D.R. is a founder and member of the scientific advisory board of Raze Thera- purine and thymidine synthesis. However, the carried peutics. G.S.D., J.M.G., H.K., and J.D.R. are inventors on a Princeton University patent through this enzyme is insufficient to meet glycine demand, and covering serine hydroxymethyltransferse inhibitors and their use in cancer. mitochondrial folate-mutant cell lines are glycine auxotrophs This article is a PNAS Direct Submission. C.V.D. is a guest editor invited by the (14). Because glycine is abundant in serum, such has Editorial Board. not been considered physiologically relevant in mammals. How- Data deposition: The atomic coordinates and factors have been deposited in the ever, recent work has identified functional amino acid shortages Data Bank, www.wwpdb.org (PDB ID 5V7I). in human tumors, suggesting that transport from serum to tumor 1To whom correspondence should be addressed. Email: [email protected]. may be limiting in some contexts, resulting in dependence on This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. intracellular synthesis (15). 1073/pnas.1706617114/-/DCSupplemental.

11404–11409 | PNAS | October 24, 2017 | vol. 114 | no. 43 www.pnas.org/cgi/doi/10.1073/pnas.1706617114 Downloaded by guest on September 26, 2021 from other cell lineages, or whether differential requirements A PHGDH mitochondria glucose 3-PG serine serine might exist for tumorigenesis versus maintenance.

SHMT1 THF SHMT2 Small-Molecule Inhibitors of Human SHMT1/2. Compounds with the pyrazolopyran scaffold represented by compound 1 (Fig. 2A)were glycine glycine described as inhibitors of plant SHMT and showed efficacy as (20). Derivatives with a meta-thiophene substitution dTMP 5,10-meTHF 5,10-meTHF were recently published as inhibitors of Plasmodium SHMT (21). MTHFD1 MTHFD2 When these compounds were tested in human , potency was poor (22). We optimized compounds of this class for human purines 10-formylTHF 10-formylTHF SHMT1 and 2 (23). Compounds of this class were modestly more potent in vitro against SHMT1 than SHMT2. Changes that improve MTHFD1 MTHFD1L potency against both human isoforms include introduction of an iso- propyl group at the chiral four-carbon of the pyrano ring and adding formate steric bulk to the metasubstitutions on the phenyl ring (compound 2). BC ) ) 3 3 HCT-116 WT * Δshmt2 Δshmt2 Δshmt1/Δshmt2 * A * Plant SHMT inhibitor Optimized human SHMT inhibitors *

Cl N OH Cl CF tumor volume (mm tumor volume (mm 3 days post injection days post injection CN CN CN Fig. 1. SHMT is required for tumor formation in vivo. (A) Serine synthesis and N N N catabolism occur in an intercompartmental cycle mediated by cytosolic and N N N O NH2 O NH2 O NH2 mitochondrial SHMT activity. Key enzymes mediating these transformations H H H are highlighted in capital letters. Arrows indicate the directionality of flux in (±)-1 (±)-2 (±)-3 HCT-116 cells, but most reactions are readily reversible. (B) Growth of sub- IC (nM) SHIN1 cutaneous tumors from HCT-116 WT and ΔSHMT2 cells implanted in opposite 50 flanks of nude mice (mean ± SEM, n = 10, *P < 0.05, paired t test). (C)Tumor SHMT1 593 13 5 growth of subcutaneous tumors from HCT-116 ΔSHMT2 and ΔSHMT1/2 cells SHMT2 1193 66 13 implanted in opposite flanks of nude mice (mean ± SEM, n = 10). B K409 Y105 take up glycine, which was previously viewed as a nonessential byproduct of the SHMT reaction. Results Y106 Requirement for SHMT Activity in HCT-116 Xenograft Formation. We generated clonal deletion cell lines of SHMT1, SHMT2, and SHMT1/ 2 from the human colorectal carcinoma cell line HCT-116. Paired GLY N410 Cas9 nickase (Cas9n)-containing constructs that encoded single-guide L166 RNA sequences targeting SHMT1 or -2 were transiently transfected 3.80Å into cells, and mutant colonies fromsinglecloneswerepickedas 2.99Å previously described (11). As previously reported, SHMT1 deletion H171 hadnoeffectoncellgrowtheitherin cell culture or as subcutaneous SHMT2/ (2) SHMT1/ (5CHO-THF) xenografts in nude mice. In contrast, SHMT2 deletion cells grew C D slower in culture and as xenografts (Fig. 1B and Fig. S1A). Liquid HCT-116 WT Δshmt1 Cellular IC (nM) -mass spectrometry (LC-MS) analysis of the soluble WT + formate Δshmt2 50 metabolites extracted from SHMT2 deletion tumors revealed char- (+)-2 (+)-SHIN1 B acteristic signs of defective serine catabolism (Fig. S1 ): serine levels WT 2300 870 were increased ∼twofold and the purine intermediate amino- +formate 13500 >50000 carboxamide ribotide (AICAR), whose consumption Δshmt1 requires 10-formyl–THF, was elevated ∼25-fold. 2800 840 Δshmt2 To generate dual SHMT1/SHMT2 double-deletion cell lines, 36 10 cell growth (norm) SHMT2 deletion cells were transfected with Cas9 and guide RNA 011 10 102 103 04 sequences targeting SHMT1 in the presence of 1 mM sodium [(+)-SHIN1] nM formate. Isolated clones cultured in formate grew at rates com-

parable to WT parental cells; no growth was observed in media Fig. 2. A folate-competitive cell-permeable inhibitor of human SHMT1/2. (A) without formate (Fig. S1 C and D). To test whether circulating Structure of pyrazolopyran inhibitor of plant SHMT (compound 1) and two opti- = and 1C sources in vivo could support the growth of mized inhibitors for human SHMT1/2 (compounds 2 and 3; compound 3 SHIN1). SHMT1/SHMT2 double-deletion cells, we xenografted them into IC50s shown are for human SHMT1 and -2 in an in vitro assay. (B, Left) Compound 2 in complex with human SHMT2 as solved in a 2.5-Å resolution X-ray crystal structure. nude mice. No tumors were observed from the SHMT1/SHMT2 – C E The electron density of the compound is shown as the 2Fo Fc mapcontouredat double-deletion cells (Fig. 1 and Fig. S1 ). Thus, in HCT-116 0.5 σ and generated with compound 2 omitted. (Right) An overlay of the SHMT2/ xenografts, circulating alternative 1C donors (e.g., , sarco- compound 2 structure with the structure of 5-formyl–THF-triglutamate in complex sine, formate) and nucleotides are together insufficient to support with rabbit SHMT1. (C)GrowthofHCT-116WT±1mMformateandΔSHMT1 and intracellular 1C metabolism required for tumorigenesis. It remains ΔSHMT2 cell lines in the presence of increasing concentrations of SHIN1 (n ≥ 3). (D)

to be tested whether SHMT activity is essential for tumors derived Cellular IC50 values for growth inhibition by compound 2 and SHIN1.

Ducker et al. PNAS | October 24, 2017 | vol. 114 | no. 43 | 11405 Downloaded by guest on September 26, 2021 Aromatic substitution at this position further increased potency, As compound 2 and SHIN1 both have similar biochemical ac- yielding compound 3, which inhibits proliferation (24). We term tivities against SHMT1 and SHMT2, the much higher doses re- this inhibitor serine hydroxymethyltranferase inhibitor 1, or SHIN1. quired for functional inhibition of cellular SHMT2 likely reflects a To understand the binding mode of these inhibitors, we solved a combination of imperfect mitochondrial penetration and greater 2.47-Å structure of human SHMT2 as a dimer in complex with intrinsic cellular SHMT2 activity (i.e., a substantial functional re- glycine, pyridoxal 5′-phosphate (PLP), and racemic compound 2 serve due to high SHMT2 expression). Importantly, the effects on (Fig. 2B and Table S1) (PDB ID code 5V7I). Electron density was cell growth of compound 2 and SHIN1 could be rescued by addition identified in both binding pockets of the , but in only of formate, indicating that they inhibit cell growth through on-target one was it well resolved. Similar to the solved structure depletion of cellular 1C pools (Fig. 2D). However, because glycine of a pyrazolopyran inhibitor in complex with Plasmodium vivax is also a product of the SHMT reaction, formate can only rescue cell SHMT (21), binding contacts with the exocyclic growth when this amino acid is present in the media (Fig. S2C). are made with the backbone of L166 and between the Notably, while most cancers have high mitochondrial 1C pathway pyrazole and H171. Overlaying our inhibitor-bound structure with activity, certain cancer cells, such as the pancreatic cancer cell line a previously solved structure of rabbit SHMT1 bound to 5-formyl– 8988T, harbor genetic lesions in the mitochondrial folate pathway THF triglutamate (PDB ID code 1LS3) revealed that the bicyclic activity and therefore rely on SHMT1 to generate 1C units (11). In 2 such cells, SHIN1 impairs cell growth at concentrations <100 nm ring system of compound and pteridine of folate occupy D thesamespace,butatadifferentangle(Fig.2B). However, hydro- due to its potent engagement of cellular SHMT1 (Fig. S2 ). gen bond contacts are preserved, and engage the inhibitor at several SHMT Target Engagement. Inhibition of cellular SHMT activity can core positions, including the exocyclic amine and the pyrazole ni- 13 trogens. The substituted phenyl ring and associated of be monitored by isotope tracers and LC-MS. U- C serine is ca- tabolized in the mitochondria by SHMT2 into U-13C-glycine and a compound 2 trace along the para-aminobenzoic acid moiety of folate 13C-5,10-methylene–THF. Glycine is further incorporated into as it exits the pteridine binding pocket toward the solvent-exposed downstream metabolites, such as and purines, whereas folate polyglutamate . Directly adjacent to the pyrrolidine π the folate 1C unit can be exported to the for incorporation lies a that is well positioned to form a -stacking into purines and thymidine. In addition, glycine and a 1C unit can interaction with the phenyl of SHIN1, potentially contributing the recombine to make partially labeled serine via SHMT1 or SHMT2 improved potency of this compound. Given the conserved nature of (Fig. S3A). To assess target engagement, we compared the effects the SHMT active site, these compounds are likely to inhibit SHMT of SHMT genetic manipulations to pharmacological treatment enzymes not only of humans, but also other mammals. with SHIN1. Serine media consumption was inhibited in both Both compound 2 and SHIN1 contain a single chiral center. Al- 2 HCT-116 SHMT1/2 double-deletion cells and WT cells treated though crystallization was performed with racemic compound ,the with (+)-SHIN1 (Fig. S3B). Glycine production from serine and electron density was consistent with only a single binding subsequent incorporation into glutathione or ADP was completely to the enzyme. Using chiral chromatography, we separated compound blocked in SHMT1/2 double-deletion cells, as evidenced by the 2 A and confirmed enantioselective enzyme inhibition (Fig. S2 ). missing M+2labelingfraction(Fig.3A). Nearly complete block- ade was observed in WT cells treated with (+)-SHIN1 but not the Cell Growth Inhibition. We next sought to investigate the activity of inactive enantiomer (−)-SHIN1. treatment also blocked 2 compound and SHIN1 against cytosolic and mitochondrial SHMT recombination of glycine and 10-formyl–THF to reform serine − isoforms in cultured cells. The inactive ( ) enantiomer of SHIN1 (Fig. S3B). Genetic deletion of SHMT1/2, and to a lesser extent had no significant effect on growth in HCT-116 cells at doses up to SHMT2, results in a build-up of purine biosynthetic intermediates 30 μM(Fig. S2B), whereas the active (+) enantiomer blocked upstream of steps requiring 10-formyl–THF as a substrate (Fig. growth with half-maximal inhibitory constants (IC50)of870nM(Fig. 3B). Such build-up is also seen with (+)-SHIN1. Thus, SHIN1 2 C and D). To analyze the effects of inhibition on each isoform inde- phenocopies—in an enantioselective manner—the metabolic con- pendently, we used the SHMT1 and SHMT2 HCT-116 deletion sequences of SHMT genetic deletion. clones. The active of both compounds, (+)-2 and To assess the selectivity of the metabolic effects of SHIN1, we (+)-SHIN1, were potent against cytosolic SHMT1, as evidenced performed untargeted LC-MS analysis on soluble metabolites by IC50 for growth of less than 50 nM in SHMT2 deletion cells (Fig. from drug-treated cells (Fig. 3C). In addition to purine interme- 2 C and D). In contrast, SHMT1 deletion cells showed indis- diates, we saw build-up of purine salvage products (xanthosine, tinguishable sensitivity from WT, confirming that mitochondrial guanosine), whose increase is consistent with purine insufficiency. SHMT inhibition is limiting for compound efficacy (Fig. 2D). We further saw build-up of , a classic marker of 1C

U13C serine tracer soluble metabolites soluble metabolites A B HCT-116 ∆shmt1/2 C D ∆shmt1 HCT-116 WT cells HCT-116 WT cells ADP (-)-SHIN1 ∆shmt2 Glutathione (+)-SHIN1 R2 = .87 R2 = .94 1

1: AICAR 3 4 6 2: SAICAR 2 5 3: xanthosine 4: guanosine TIC (normalized) 5: homocysteine (+)-SHIN1 (TIC) 6: n-carbamoyl asp GAR FGAR SAICAR AICAR M+2 labeling frac. +gly shmt2 (+)-SHIN1 + 1mM F (TIC) ∆shmt1∆ shmt1/2 HCT-116 ∆ (-)-SHIN1 +10-formylTHF +10-formylTHF (+)-SHIN1 DMSO (TIC) DMSO (TIC)

Fig. 3. (+)-SHIN1 inhibits SHMT1/2 in HCT-116 cells. (A)M+2 13C-labeling fraction of intracellular ADP and glutathione after 24 h 13C-serine coincubation with DMSO, 5 μM(+)-SHIN1, or 5 μM(–)-SHIN1. ΔSHMT1/2 cells were cultured without formate for the duration of labeling (mean ± SD, n = 3). (B) Normalized (to DMSO HCT-116 WT cells) levels of purine biosynthetic pathway intermediates after 24-h incubation ±SHIN1 (mean ± SD, n = 3). (C) Total metabolite abundances in HCT-116 cells treated with DMSO vs. (+)-SHIN1 (10 μM) for 48 h. Metabolites whose abundances differ by more than fourfold between conditions are highlighted in red (mean, n = 3). (D) Metabolite abundance in HCT-116 cells treated with DMSO or SHIN1 in the presence of 1 mM sodium formate. The same metabolites whose abundances were different in C are highlighted in red (mean, n = 3).

11406 | www.pnas.org/cgi/doi/10.1073/pnas.1706617114 Ducker et al. Downloaded by guest on September 26, 2021 A B media C Su-DHL-4 <4 μM ≥4 μM + 1mM formate Jurkat B cell 24 8 other 120 146 ***

) (+)- 2 p<.01, Fisher’s exact test 50 ) (+)-SHIN1

50 * *** *** -log(IC -log(IC DLBCL AV+/PI- fraction RL

Burkitt’s JM1 Daudi JVM-2 ST486 MC116 MM.1S OPM-2 Farage MM.1R HuNS1 SUP-T1 ARH-77 Other U266B1 DOHH-2 (+)-SHIN1 + + + + SUP-B15 -- NCI-H929 SU-DHL-4 SU-DHL-2 SU-DHL-6 NAMALWA Jurkat E6-1 Ramos-2G6 + + 298 cell lines WSU-DLCL2 formate -- -- ALL MM Burkitt’s DLBCL

Fig. 4. SHMT inhibitors are particularly active against B-cell malignancies. (A) Ranked IC50, in units of molarity, of compound (+)-2 for growth inhibition of 298 human cancer cell lines. Lines of B-cell origin are highlighted in red and are enriched among the more sensitive cells (IC50 < 4 μM). (B)IC50 of (+)-SHIN1, with and without 1 mM formate, for growth inhibition of selecthematologicalcelllines.(C) Fraction of Jurkat and Su-DHL-4 cells that are apoptotic after 24 h (+)-SHIN1 treatment (2.5 and 5 μM respectively) as indicated by flow cytometry using FITC-Annexin V staining (mean ± SD, n ≥ 3, *P < 0.05, ***P < 0.001, unpaired t test).

deficiency. We also observed depletion of the pyrimidine in- DLBCL cell lines suggested that glycine may be limiting in these termediate N-carbamoyl-aspartate, likely reflecting feedback in- cells. To explore this hypothesis, we characterized the metabolic ef- hibition of aspartate transcarbamoylase by excess pyrimidines in fects of SHIN1 in DLBCL and Jurkat cells treated with (+)-SHIN1 the purine-starved cells (25). Importantly, there were no other large (72 h, 5 μM) with and without formate. In Jurkat and DLBCL cell changes in metabolism, suggestive of off-target effects. Moreover, lines Su-DHL-4 and Su-DHL-2 in the absence of formate, SHIN1 the changes in abundances were rescued by formate (Fig. 3D)and treatment led to a large reduction in nucleotide triphosphates (Fig. metabolite abundances in SHMT1/2 double-deletion cells closely 5A and Fig. S5A). This can be rationalized as reflecting impaired matched those from WT cells treated with (+)-SHIN1 (Fig. S3C). purine synthesis, which requires both 1C units and glycine, with Thus, at doses sufficient to robustly inhibit SHMT1 and SHMT2 in pyrimidines also falling due to endogenous mechanisms that cell culture, (+)-SHIN1 selectively targets 1C metabolism. balance their levels with those of purines. There is also a com- ponent of energy stress, particularly in Su-DHL-4 cells, as nucle- Cancer Cell Line Sensitivity to SHMT Inhibition. Unfortunately, SHIN1 otide monophosphates were increased, not decreased (Fig. S5B). and related pyrazolopyrans are unstable in microsome assays Consistent with 1C limitation, dTTP, whose synthesis requires a and have poor in vivo half-, precluding their immediate use in folate 1C unit, was more depleted than other pyrimidines. animal models. Accordingly, we focused on their in vitro applica- Formate supplementation restored nucleotide levels in Jurkat tion across a wide range of cancer cell lines. Specifically, we but not DLBCL cell lines. We confirmed that formate rescues screened a panel of nearly 300 human cancer cell lines for growth folate 1C levels in DLBCL cells, as the AICAR accumulation inthepresenceofthe(+)-enantiomer of compound 2 (Fig. 4A and induced by (+)-SHIN1 is fully reversed (Fig. S5C). Thus, while Table S2). The median IC50 was 4 μM. Cell lines of B-cell lymphoma nucleotide synthesis in SHIN1-treated Jurkat cells is solely limited origin were enriched in the more sensitive half of cells (P < 0.001, by 1C units, an additional factor is lacking in DLBCL cells. Fisher’s exact test). This effect was driven by a pronounced sensitivity Consistent with glycine being the second factor missing in DLBCL of Burkitt’s and DLBCL lymphomas (Fig. 4A). We then rescreened a cells, (+)-SHIN1 treatment depleted the glycine-containing redox set of hematological cancer lines with (+)-SHIN1, supplemented defense glutathione (Fig. 5B and Fig. S5D). Strikingly, with and without formate to test for rescue (Fig. 4B). Like HCT-116 while SHIN1 alone did not alter glutathione in Jurkat cells, for- cells, cell lines of T cell origin, such as acute lymphocytic leukemia mate addition caused glutathione depletion. This further validates (ALL) cells, were largely rescued from the antigrowth effects of that, when SHMT is inhibited, provision of excess 1C units can (+)-SHIN1 by formate (Fig. 4B, gray bars). In contrast, formate cause glycine stress. Glutathione supplementation did not rescue failed to rescue the growth of B-cell lymphoma lines. growth (Fig. S5E). Based on these results, we predicted that To explore this surprising lack of rescue further, we analyzed growth in SHMT-inhibited DLBCL cells might be restored with by flow cytometry the effect of (+)-SHIN1, with and without purine supplementation, which would simultaneously alleviate 1C formate, on the DLBCL cell line Su-DHL-4. SHIN1 itself in- and glycine metabolic stress. Growth was partially rescued in Su- duced apoptosis as measured by Annexin V surface staining (Fig. DHL-4 cells treated with hypoxanthine (Fig. 5C). Thymidine, 4C and Fig. S4A). Apoptosis was enhanced by cotreatment with which rescues the effects of the classic antifolate pemetrexed but formate. In contrast, as expected, formate rescued Jurkat E6-1 does not contain glycine, had no benefit in SHIN1-treated DLBCL leukemia cells from apoptosis (Fig. 4C and Fig. S4B). cells. Thus, SHIN1 blocks cell growth through a progressive To account for these observations, we hypothesize that the depletion of purines, leading to loss of nucleotide triphosphates. failure of formate to rescue growth in the DLBCL cell lines is Restoration of purines levels restores growth. due to a requirement for both glycine and 1C units made by The depletion of glycine-derived metabolites in DLBCL cells SHMT in these cells. When glycine is limiting, formate can en- led us to examine whether glycine shortage might also impact hance the cytotoxicity of SHMT inhibition. For example, formate protein synthesis. Severe amino acid shortages lead to loss of augments the effect of SHIN1 in HCT-116 cells in glycine-free cognate tRNA charging and thus stalling, which can be BIOCHEMISTRY media (Fig. S2C). Mechanistically, by supplying 5,10-methylene– measured using ribosome profiling (15). We performed ribosome THF, formate may drive residual SHMT enzymatic function in profiling on Su-DHL-4 cells treated with (+)-SHIN1 (Fig. S6 A the glycine-consuming direction. Alternatively, whereas cells and B). Untreated Su-DHL-4 cells growing in RPMI did not show may have the machinery to sense 1C deficiency and safely pause evidence of glycyl-tRNA insufficiency; no enrichment for these growth (e.g., due to AICAR activation of AMPK), they may lack codons was observed (Fig. S6C). Furthermore, we did not observe comparable mechanisms for surviving glycine limitation. any difference in glycine codon occupancy between treated and control cells (Fig. S6D). Collectively, these results suggest a DLBCL Cells Require SHMT to Make Glycine for Purine Synthesis. The hierarchy in the sensitivity of different intracellular metabolic inability of formate to rescue the antigrowth effects of SHIN1 in products to glycine levels: glutathione synthesis is most sensitive,

Ducker et al. PNAS | October 24, 2017 | vol. 114 | no. 43 | 11407 Downloaded by guest on September 26, 2021 followed by purine synthesis, with protein synthesis most resistant. A Jurkat Su-DHL-4 B This hierarchy is consistent with biochemical measurements of the Jurkat K ATP Su-DHL-4 m values of the relevant enzymes: the glycyl-tRNA amino acid GTP synthase has a lower Km for glycine (15 μM) than that found in CTP ribonucleotide synthetase (45 μM) or glutathione UTP synthetase (452 μM) (26–28). dTTP Defective Glycine Uptake in DLBCL. SHIN1 induced glycine deficiency in DLBCL cells, even though they were cultured in complete media with glycine (RPMI, 10 mg/L glycine = 130 μM). This glutathione (norm.) total count (norm.) suggested that glycine uptake is intrinsically impaired in these cells. 13 + + (+)-SHIN1 - - + (+)-SHIN1 - Using U- C-glycine, we monitored the kinetics of extracellular ++

++ formate glycine incorporation into cells and downstream metabolic prod- formate -- -- ++ -- ucts (Fig. 5D). Labeling of intracellular glycine products, such as U13C glycine tracer C DMSO D glutathione and ADP, was markedly less in Su-DHL-4 cells than

) gly 5 (+)-SHIN1 Jurkat Jurkat cells. In a larger set of cell lines, composed of both other SHIN1+hypo GSH ADP hematological cancer and adherent cell lines, steady-state labeling SHIN1+thym of intracellular metabolites from glycine was significantly lower in B-cell lymphoma cell lines (Fig. 5E). Su-DHL-4 Given the apparent glycine shortage in these B cells upon SHIN1 treatment, we next sought to augment extracellular glycine Su-DHL-4

M+2 labeling frac. levels and evaluate response to drug. We first altered the concen- cell number (1x10 trationofglycineinRPMIandobservedresponsetodrug.Are- duction of glycine in the media modestly improved the potency of time (hrs) time (hrs) SHIN1, indicating that the cells were sensitive to extracellular glycine. E purine (ADP) glutathione More strikingly, increasing the media glycine by 10-fold substantially adherent LN-229 rescued the cells from SHIN1 (Fig. S5F). In contrast, in Jurkat cells, a HCT-116 small amount of extracellular glycine was sufficient and more did suspension REH not further rescue the cells from SHIN1 (Fig. S5G). Across a set of Jurkat DLBCL cell lines, representing both ABC and GBC subtypes, sup- B cell Su-DHL-2 plying both formate and supraphysiologic glycine (100 mg/L, 1.3 mM) Farage generally rescued cell growth (Fig. 5F). These results indicate the Daudi importance of both products of the SHMT reaction, glycine and fo- Su-DHL-4 late 1C units, for the proliferation of DLBCL cell lines. Knowing that manipulating glycine could augment the efficacy M+2 labeling fraction (U13C glycine tracer) of SHIN1, we tested different mechanisms to decrease glycine. As observed previously, when formate was added, SHIN1 was trans- F DLBCL formed from being a drug that slowed cell growth to one that was fully cytostatic (Fig. 5G). Further removing glycine caused signif- icant cell death. Interestingly, combining the glycine transporter 1 (GlyT1; SLC6A9) inhibitor RG1678 with SHIN1 further increased cell death, even in the presence of media glycine (29) (Fig. 5G). These results suggest that glycine uptake in these

cell grwoth (norm) cells is mediated by GLYT1 and that combinations of formate, SHMT inhibitor, and GLYT1 inhibitor may selectively target 2.5 μM (+)-SHIN1 - +++ - +++ - +++ - +++ - +++ - +++ these cells. 1 mM formate -- ++ -- ++ -- ++ -- ++ -- ++ -- ++ 100 mg/L (10X) gly --- + --- + --- + --- + --- + --- + Discussion REHJurkat Su-2 Su-4 Farage Su-6 Targeting folate metabolism has been employed clinically to treat cancer for over 70 y (30). Despite the use of antifolates and other G Su-DHL-4 H antimetabolites in many important chemotherapy regimens, their ) 2 SHMT inhibition in B cells clinical effectiveness is limited by side effects in normal pro- liferating . Identifying metabolic processes that can be tar- serine serine geted in a more tumor-selective manner remains a major challenge.

SHMT SHMT SHIN1

fold change in Jurkat and Su-DHL-4 (gly, glycine; GSH, glutathione; mean ± SD, n = 3). (E) cell number (log meTHF glycine The steady-state labeling fraction of intracellular metabolites synthesized meTHF glycine 13 + (+)-SHIN1 - ++ - from glycine in cancer cell lines cultured in RPMI containing U- C-glycine (mean ± SD, n = 3). (F) Cell growth (normalized to DMSO) of DLBCL and + + ++ glycine - + μ ++ + + formate - - + - other hematopoietic cancer lines with 2.5 M SHIN1, in RPMI with or without glycine formate glycine × RG1678 -- -- + 1 mM formate and 10 physiological glycine (100 mg/L); all conditions in- cluded at least normal media glycine (10 mg/L; mean ± SD, n = 3). (G) Cell

Fig. 5. Glycine made by SHMT is required for B-lymphoma cell line growth. growth (or death) as measured by log2-fold change in cell number over 48 h (A) Normalized total ion counts of nucleotide triphosphates in Jurkat ALL in Su-DHL-4 cells cultured in RPMI with and without glycine (10 mg/L), for- cells and Su-DHL-4 DLBCL cells after 72-h treatment with (+)-SHIN1 (5 μM). mate (1 mM), the glycine transporter inhibitor RG1678 (300 nM), and/or Coculture with 1 mM formate restores nucleotide levels selectively in Jurkat (+)-SHIN1 (5 μM) (mean ± SD, n = 3). (H) Schematic illustrating the proposed cells (mean ± SD, n = 3–6). (B) Normalized glutathione levels from Jurkat and glycine vulnerability in B cells. The SHMT reaction makes two products, 5,10- Su-DHL-4 cells treated as in A (mean ± SD, n = 3–6). (C) Growth of Su-DHL- methylene–THF and glycine. When SHMT is inhibited, exogenous formate 4 cells treated with (+)-SHIN1 and hypoxanthine (100 μM) or thymidine (16 μM; can be incorporated into the 1C cycle, whereas in B cells poor glycine uptake mean ± SD, n = 3). (D) Intracellular U-13C-glycine assimilation kinetics in limits the ability of extracellular glycine to rescue.

11408 | www.pnas.org/cgi/doi/10.1073/pnas.1706617114 Ducker et al. Downloaded by guest on September 26, 2021 In this study, we targeted the SHMT reaction, which uses serine synthesis in ALL, which creates a dependence upon external to generate a folate-bound 1C unit and glycine. Consistent with sources of (31). This dependence is targeted by aspar- prior reports (22), we found that pyrazolopyrans have detectable aginase, a core medicine in pediatric ALL therapy (32). In DLBCL, activity against human SHMT. Through substantial chemistry ef- because the defect is in glycine transport rather than synthesis, the forts, we enhanced the potency for human SHMT by over 100-fold, therapeutic strategy rests on inhibiting intracellular glycine synthe- resulting in inhibitors such as SHIN1 with on-target dual SHMT1/ sis. The resulting efficacy can be increased either by decreasing 2 cellular inhibition at nanomolar to low micromolar concentrations. extracellular glycine, or more promisingly therapeutically, by further We extensively validated the on-target activity of these compounds depleting intracellular glycine by formate addition or glycine uptake using metabolomics in combination with genetics. While these com- inhibition. Formate is attractive because it can rescue the effects of pounds have appropriate stability for cell culture studies, including SHMT inhibition in normal tissues with strong glycine uptake. Both of primary T cells (24), they are not currently usable in vivo due to approaches however, may exacerbate toxicity in tissues with natu- rapid clearance. rally low glycine transport. While glycine transport is poorly char- Screening of cancer cell lines for sensitivity to small-molecule acterized in vivo in most tissues, existing data suggest immune and SHMT1/2 inhibitors revealed specific metabolic vulnerabilities of certain cancers. One mode of sensitization, exemplified by the pan- neurological tissues may be potentially sensitive to modulation of creatic cancer cell line 8988T, results from defects in mitochondrial glycine synthesis. Going forward, a careful assessment of amino acid folate metabolism. Such cells are dependent upon SHMT1 for pro- transport in vivo will be required to understand how to best exploit duction of 1C units, and functionally have low reserve SHMT ac- glycine transport defects for therapy. tivity, rendering them sensitive to low concentrations of SHIN1. By a different mechanism, B-cell lymphomas are also uniquely Methods sensitive to SHMT inhibition. We show that these cells are intrin- All mouse work was approved by the Princeton University Institutional Animal sically deficient in glycine uptake and thus require glycine made by Care and Use Committee. For metabolite measurements, cultured cells were SHMT to grow. When combined with formate, SHMT inhibitors do incubated in media containing dialyzed FBS and the isotopically labeled me- not function as classic antifolates by disrupting 1C metabolism, but tabolite of interest. Cells were quenched with cold methanol and metabolites rather, in cells with impaired glycine uptake, as cell-type–specific analyzed by LCMS. Full-length human SHMT1 and -2 protein was isolated from glycine depletion agents (Fig. 5H). As a fundamental precursor to using nickel capture followed by cleavage of the HIS tag using many essential , glycine is in high demand. Indeed, the tobacco etch virus (TEV) . Complete chemical synthesis details and quantitative demand for glycine to support protein, nucleotide, and compound characterizations are provided in Chemical Synthesis Methods.All glutathione synthesis exceeds the cellular requirement for 1C units experimental procedures are described in detail in SI Methods. (11). Using metabolomics and ribosome profiling, we characterized the susceptibility of these processes to glycine stress. Consistent with ACKNOWLEDGMENTS. We thank C. DeCoste of the Princeton University flow the reported enzymatic K values, glutathione and purine synthesis cytometry resource facility for experimental set-up and design; R. Morscher for m assistance with animal experiments; and members of the J.D.R. laboratory for were more sensitive than protein synthesis to glycine depletion. general assistance and discussions. G.S.D. received past fellowship support from Targeting an amino acid vulnerability is a well-established thera- the American Cancer Society (PF-15-190-01-TBE), and is currently supported by peutic strategy in cancer. A useful comparison with the intrinsic NIH Award K99CA215307. J.D.R. is supported by Grant SU2CAACR-DT-20- defect in glycine uptake in DLBCL is the defect in asparagine 16 from Stand Up 2 Cancer and NIH Award R01CA163591.

1. Pavlova NN, Thompson CB (2016) The emerging hallmarks of cancer metabolism. Cell 20. Witschel M, Stelzer F, Hutzler J, Qu T (2013) Pyrazolopyrans having herbicidal and Metab 23:27–47. pharmaceutical properties. US Patent 61/656,025. 2. Schulze A, Harris AL (2012) How cancer metabolism is tuned for proliferation and 21. Witschel MC, et al. (2015) Inhibitors of plasmodial serine hydroxymethyltransferase vulnerable to disruption. Nature 491:364–373. (SHMT): Cocrystal of pyrazolopyrans with potent - and liver-stage 3. Labuschagne CF, van den Broek NJF, Mackay GM, Vousden KH, Maddocks ODK (2014) activities. J Med Chem 58:3117–3130. Serine, but not glycine, supports one-carbon metabolism and proliferation of cancer 22. Marani M, et al. (2016) A pyrazolopyran derivative preferentially inhibits the activity cells. Cell Rep 7:1248–1258. of human cytosolic serine hydroxymethyltransferase and induces cell death in lung 4. Locasale JW, et al. (2011) Phosphoglycerate dehydrogenase diverts glycolytic flux and cancer cells. Oncotarget 7:4570–4583. contributes to oncogenesis. Nat Genet 43:869–874. 23. Rabinowitz JD, Kim H, Ducker GS, Ghergurovich JM (2016) Patent Cooperation Treaty 5. Possemato R, et al. (2011) Functional genomics reveal that the serine synthesis Appl WO2016145252A1 (September 15, 2016). pathway is essential in breast cancer. Nature 476:346–350. 24. Ma EH, et al. (2017) Serine is an essential metabolite for effector T cell expansion. Cell 6. Jain M, et al. (2012) Metabolite profiling identifies a key role for glycine in rapid – cancer cell proliferation. Science 336:1040–1044. Metab 25:345 357. 7. Lee GY, et al. (2014) Comparative oncogenomics identifies PSMB4 and SHMT2 as 25. Cockrell GM, et al. (2013) New paradigm for of Escherichia coli – potential cancer driver . Cancer Res 74:3114–3126. aspartate transcarbamoylase. Biochemistry 52:8036 8047. 8. Nilsson R, et al. (2014) Metabolic enzyme expression highlights a key role for 26. Cader MZ, et al. (2007) Crystal structure of human wildtype and S581L-mutant glycyl-tRNA MTHFD2 and the mitochondrial folate pathway in cancer. Nat Commun 5:3128. synthetase, an enzyme underlying distal spinal muscular atrophy. FEBS Lett 581:2959–2964. 9. Ben-Sahra I, Hoxhaj G, Ricoult SJH, Asara JM, Manning BD (2016) mTORC1 induces purine 27. Antle VD, et al. (1996) Substrate specificity of glycinamide ribonucleotide synthetase synthesis through control of the mitochondrial tetrahydrofolate cycle. Science 351:728–733. from chicken liver. J Biol Chem 271:8192–8195. 10. Lewis CA, et al. (2014) Tracing compartmentalized NADPH metabolism in the cytosol 28. Njalsson R, et al. (2001) Cooperative binding of γ-glutamyl substrate to human glu- and mitochondria of mammalian cells. Mol Cell 55:253–263. tathione synthetase. Biochem Biophys Res Commun 289:80–84. 11. Ducker GS, et al. (2016) Reversal of cytosolic one-carbon flux compensates for loss of 29. Harvey RJ, Yee BK (2013) Glycine transporters as novel therapeutic targets in the mitochondrial folate pathway. Cell Metab 23:1140–1153. schizophrenia, alcohol dependence and pain. Nat Rev Drug Discov 12:866–885. 12. Tibbetts AS, Appling DR (2010) Compartmentalization of mammalian folate-medi- 30. Farber S, Diamond LK, Mercer RD, Sylvester RF, Wolff JA (1948) Temporary remissions – ated one-carbon metabolism. Annu Rev Nutr 30:57 81. in acute leukemia in children produced by folic acid antagonist, 4-aminopteroyl- 13. Patel H, Pietro ED, MacKenzie RE (2003) Mammalian fibroblasts lacking mitochondrial (aminopterin). N Engl J Med 238:787–793. + NAD -dependent methylenetetrahydrofolate dehydrogenase-cyclohydrolase are 31. Kiriyama Y, et al. (1989) Biochemical characterization of U937 cells resistant to glycine auxotrophs. J Biol Chem 278:19436–19441. L-: The role of . Leukemia 3:294–297.

14. Ducker GS, Rabinowitz JD (2017) One-carbon metabolism in health and disease. Cell BIOCHEMISTRY 32. Pui C-H, Evans WE (2006) Treatment of acute lymphoblastic leukemia. N Engl J Med Metab 25:27–42. 354:166–178. 15. Loayza-Puch F, et al. (2016) Tumour-specific vulnerability uncovered by dif- 33. Ran FA, et al. (2013) Double nicking by RNA-guided CRISPR Cas9 for enhanced ge- ferential ribosome codon reading. Nature 530:490–494. nome editing specificity. Cell 154:1380–1389. 16. Zhao R, Goldman ID (2003) Resistance to antifolates. Oncogene 22:7431–7457. 34. Giardina G, et al. (2015) How pyridoxal 5′-phosphate differentially regulates human cytosolic 17. Pacold ME, et al. (2016) A PHGDH inhibitor reveals coordination of serine synthesis – and one-carbon unit fate. Nat Chem Biol 12:452–458. and mitochondrial serine hydroxymethyltransferase oligomeric state. FEBS J 282:1225 1241. 18. Wang Q, et al. (2017) Rational design of selective allosteric inhibitors of PHGDH and 35. Lu W, et al. (2010) Metabolomic analysis via reversed-phase ion-pairing liquid chro- serine synthesis with anti-tumor activity. Cell Chem Biol 24:55–65. matography coupled to a stand alone orbitrap mass spectrometer. Anal Chem 82: 19. Mullarky E, et al. (2016) Identification of a small molecule inhibitor of 3-phospho- 3212–3221. glycerate dehydrogenase to target serine in cancers. Proc Natl Acad Sci 36. Dunn JG, Weissman JS (2016) Plastid: Nucleotide-resolution analysis of next-genera- USA 113:1778–1783. tion and genomics data. BMC Genomics 17:958.

Ducker et al. PNAS | October 24, 2017 | vol. 114 | no. 43 | 11409 Downloaded by guest on September 26, 2021