This table shows purity of Subcellular Fractions.

This table shows IC50 values for antiproliferative activities toward EOC versus normal ovary cell lines by 5-substituted pyrrolo[3,2-d]pyrimidine inhibitors.

This figure shows generation of SKOV3 SHMT1 and SHMT2 knockdown (KD) cell lines.

This table shows information of ovarian cancer tissue cDNA array for real-time RT-PCR.

This figure shows results of apoptosis assays with SKOV3 cells treated with AGF347.

This table shows primer sequences for quantitative real-time RT-PCR.

This figure shows immunohistochemical staining of tumor microarray (TMA) for SHMT2 protein.

This figure shows purine nucleotides in SKOV3 EOC cells treated with pyrrolo[3,2-d]pyrimidine antifolates.

This table shows information of epithelial ovarian cancer tissue microarray (TMA) for immunohistochemistry.

This table shows Kis for inhibition of one-carbon enzymes for 5-substituted pyrrolo[3,2- d]pyrimidine antifolates.

This figure shows purification of the recombinant human thymidylate synthase (TS) and in vitro inhibition by pyrrolo[3,2-d]pyrimidine antifolates.

This figure shows FRα, PCFT and RFC expression in EOC cell line models and EOC patient cDNAs.

One-carbon (C1) metabolism includes cytosolic and mitochondrial pathways connected by interchange between serine, glycine, and formate. Mitochondrial C1 metabolism through serine hydroxymethyltransferase (SHMT) 2 generates glycine and C1 units for de novo nucleotide biosynthesis in the cytosol, whereas cytosolic SHMT1 consumes C1 units and glycine. Folates and classical antifolates are transported into tumors by facilitative folate transporters (reduced folate carrier [RFC] and proton-coupled folate transporter [PCFT]) and are metabolized to polyglutamates by folylpolyglutamate synthetase (FPGS). Folate transporter-null HeLa cells were engineered to express RFC under the control of a tetracycline-inducible promoter. Constitutive expression of PCFT and/or FPGS increased cytosolic and mitochondrial folates over that of RFC alone. By targeted metabolomics, the C1 flux in mitochondria through SHMT2 paralleled RFC transport and folate accumulation in mitochondria and cytosol, whereas the SHMT1 flux was constant. Expression of PCFT resulted in further increased C1 flux through SHMT2, in excess of SHMT1. In vitro inhibition of cell proliferation by targeting SHMT1/2 with pyrrolo[3,2-d]pyrimidine antifolates (eg, AGF347) decreased with increasing RFC and with PCFT. Inhibition by AGF347 (not SHIN1/2) was stimulated with ectopic FPGS, accompanying increased AGF347 polyglutamates; decreased sensitivities were seen for nonclassical SHMT1/2 inhibitors (SHIN1/2), which are neither substrates for facilitative transport nor polyglutamylation. Our results document the complex interrelationships among (anti)folate membrane transport, polyglutamylation, and C1 fluxes through SHMT1 and SHMT2. They also demonstrate the profound impact of physiologic folates on antitumor activities and the extraordinary promise of multitargeted pyrrolo[3,2-d]pyrimidine antifolates for cancer therapy. SIGNIFICANCE STATEMENT: Novel pyrrolo[3,2-d]pyrimidine antifolates typified by AGF347 target serine hydroxymethyltransferase (SHMT) 2 in the mitochondria and SHMT1 and de novo purine biosynthesis in the cytosol. This manuscript documents the complex interrelationships among (anti)folate membrane transport, polyglutamylation, and one-carbon fluxes through SHMT1 and SHMT2 in the context of physiologic folate levels. The results document the therapeutic promise of classical multitargeted pyrrolo[3,2-d]pyrimidine antifolates typified by AGF347. These novel compounds offer an exciting new platform for one-carbon-targeted drug development for cancer.

Abstract One-carbon (C1) metabolism is compartmentalized between the cytosol and mitochondria with the mitochondrial C1 pathway as the major source of glycine and C1 units for cellular biosynthesis. Expression of mitochondrial C1 genes including SLC25A32, serine hydroxymethyl transferase (SHMT) 2, 5,10-methylene tetrahydrofolate dehydrogenase 2, and 5,10-methylene tetrahydrofolate dehydrogenase 1-like was significantly elevated in primary epithelial ovarian cancer (EOC) specimens compared with normal ovaries. 5-Substituted pyrrolo[3,2-d]pyrimidine antifolates (AGF347, AGF359, AGF362) inhibited proliferation of cisplatin-sensitive (A2780, CaOV3, IGROV1) and cisplatin-resistant (A2780-E80, SKOV3) EOC cells. In SKOV3 and A2780-E80 cells, colony formation was inhibited. AGF347 induced apoptosis in SKOV3 cells. In IGROV1 cells, AGF347 was transported by folate receptor (FR) α. AGF347 was also transported into IGROV1 and SKOV3 cells by the proton-coupled folate transporter (SLC46A1) and the reduced folate carrier (SLC19A1). AGF347 accumulated to high levels in the cytosol and mitochondria of SKOV3 cells. By targeted metabolomics with [2,3,3–2H]L-serine, AGF347, AGF359, and AGF362 inhibited SHMT2 in the mitochondria. In the cytosol, SHMT1 and de novo purine biosynthesis (i.e., glycinamide ribonucleotide formyltransferase, 5-aminoimidazole-4-carboxamide ribonucleotide formyltransferase) were targeted; AGF359 also inhibited thymidylate synthase. Antifolate treatments of SKOV3 cells depleted cellular glycine, mitochondrial NADH and glutathione, and showed synergistic in vitro inhibition toward SKOV3 and A2780-E80 cells when combined with cisplatin. In vivo studies with subcutaneous SKOV3 EOC xenografts in SCID mice confirmed significant antitumor efficacy of AGF347. Collectively, our studies demonstrate a unique metabolic vulnerability in EOC involving mitochondrial and cytosolic C1 metabolism, which offers a promising new platform for therapy.

We previously discovered first-in-class multitargeted 5-substituted pyrrolo[3,2-d]pyrimidine antifolates that inhibit serine hydroxymethyltransferase 2 (SHMT2), resulting in potent in vitro and in vivo antitumor efficacies. In this report, we present crystallographic structures for SHMT2 in complex with an expanded series of pyrrolo[3,2-d]pyrimidine compounds with variations in bridge length (3-5 carbons) and the side chain aromatic ring (phenyl, thiophene, fluorine-substituted phenyl, and thiophene). We evaluated structural features of the inhibitor-SHMT2 complexes and correlations to inhibitor potencies (i.e., Kis), highlighting conserved polar contacts and identifying 5-carbon bridge lengths as key determinants of inhibitor potency. Based on the analysis of SHMT2 structural data, we investigated the impact of mutation of Tyr105 in SHMT2 kinetic analysis and studies with HCT116 cells with inducible expression of wild-type and Y105F SHMT2. Increased enzyme inhibition potency by the pyrrolo[3,2-d]pyrimidine inhibitors with Phe105 SHMT2 accompanied an increased growth inhibition of Phe105-expressing HCT116 cells compared to wild-type SHMT2. Pyrrolo[3,2-d]pyrimidine inhibitors with polyglutamate modifications were evaluated for potencies against SHMT2. We determined the crystal structures of SHMT2 in complex with our lead antifolate AGF347 lacking L-glutamate, or as a diglutamate and triglutamate, for comparison with parent AGF347. These data provide the first insights into the influence of antifolate polyglutamylation on SHMT2:inhibitor interactions. Collectively, our results provide new insights into the critical structural determinants of SHMT2 binding by pyrrolo[3,2-d]pyrimidine inhibitors as novel antitumor agents, as well as the first structural characterization of human SHMT2 in complex with polyglutamates of an SHMT2-targeted antifolate.

Folate-dependent one-carbon (C1) metabolism encompasses distinct cytosolic and mitochondrial pathways connected by an interchange among serine, glycine, and formate. In both the cytosol and mitochondria, folates exist as polyglutamates, with polyglutamylation catalyzed by folylpolyglutamate synthetase (FPGS), including cytosolic and mitochondrial isoforms. Serine is metabolized by serine hydroxymethyltransferase (SHMT)2 in the mitochondria and generates glycine and C1 units for cellular biosynthesis in the cytosol. AGF347 is a novel pyrrolo[3,2-day]pyrimidine antifolate that targets SHMT2 in the mitochondria and SHMT1 and de novo purine biosynthesis in the cytosol. FPGS is expressed in primary pancreatic cancer specimens, and FPGS levels correlate with in vitro efficacies of AGF347 toward human pancreatic cancer cells. MIA PaCa-2 pancreatic cancer cells with CRISPR knockout of FPGS were engineered to express doxycycline-inducible FPGS exclusively in the cytosol (cFPGS) or in both the cytosol and mitochondria (mFPGS). Folate and AGF347 accumulations increased in both the cytosol and mitochondria with increased mFPGS but were restricted to the cytosol with cFPGS. AGF347-Glu5 inhibited SHMT2 ∼19-fold greater than AGF347. By metabolomics analysis, mFPGS stimulated the C1 flux from serine in the mitochondria and de novo purine and dTTP synthesis far greater than cFPGS. mFPGS enhanced in vitro inhibition of MIA PaCa-2 cell proliferation by AGF347 (∼30-fold) more than cFPGS (∼4.9-fold). Similar results were seen with other pyrrolo[3,2-d]pyrimidine antifolates (AGF291, AGF320); however, elevated mFPGS adversely impacted inhibition by the nonclassical SHMT2/SHMT1 inhibitor SHIN1. These results suggest a critical role of mFPGS levels in determining antitumor efficacies of mitochondrial-targeted pyrrolo[3,2-d]pyrimidine antifolates for pancreatic cancer. SIGNIFICANCE STATEMENT AGF347 is a novel pyrrolo[3,2-d]pyrimidine antifolate that targets serine hydroxymethyltransferase (SHMT)2 in the mitochondria and SHMT1 and de novo purine biosynthesis in the cytosol. AGF347 accumulation increases with folylpolyglutamate synthetase (FPGS) levels in both the cytosol and mitochondria. Increased mitochondrial FPGS stimulated one-carbon metabolic fluxes in the cytosol and mitochondria and substantially enhanced in vitro inhibition of pancreatic cancer cells by AGF347. Mitochondrial FPGS levels play important roles in determining the antitumor efficacies of pyrrolo[3,2-d]pyrimidine antifolates for pancreatic cancer.

Supplementary Figure S1 shows in vitro antiproliferative activity and targeted pathways of previously reported AICARFTase and GARFTase inhibitors. Supplementary Figure S2 shows docking of AGF291, AGF320, and AGF347 (along with 5-formyl-THF) in human SHMT2 and rabbit SHMT1. Supplementary Figure S3 shows plasma membrane folate transporter expression levels in H460, HCT116, and MIA PaCa-2 human tumor cell lines as compared to those in the IGROV-1 epithelial ovarian cancer cell line. Supplementary Figure S4 shows in vitro antiproliferative activity and targeted pathways of AGF291, AGF320, and AGF347 along with previously reported GARFTase inhibitor AGF94 in H460, HCT116, and MIA PaCa-2 cell lines. Supplementary Figure S5 shows targeted metabolomics data on total serine, serine isotope labeling patterns, total GAR, total AICAR, total adenine nucleotides, and total dTTP along with isotope labeling patterns of H460, HCT116, and MIA PaCa-2 cell lines treated with AGF291, AGF320, and AGF347 not shown in main text Figure 4. Supplementary Figure S6 shows a Western blot confirming knockdown of SHMT2 in H460 SHMT2 KD cell line and knockout of SHMT2 in HCT116 SHMT2 KO cell line. Supplementary Figure S7 shows in vivo efficacies of AGF347 and gemcitabine towards MIA PaCa-2 early and late stage tumor xenograft models where mouse serum folate levels were not depleted to approximate those found in humans. Supplementary Figure S8 shows a cytochrome c oxidase assay on tumors harvested from the metabolomics arm of the late stage in vivo AGF347 trial. Table S1 shows docking scores of novel compounds in human SHMT2 and rabbit SHMT1 (corresponding to Supplementary Figure S2). Table S2 shows quantitative data from in vivo early- and late-stage trials of AGF347 and gemcitabine against MIA PaCa-2 tumor xenografts in NCR SCID mice.

Multitargeted agents provide tumor selectivity with reduced drug resistance and dose-limiting toxicities. We previously described the multitargeted 6-substituted pyrrolo[3,2-d]pyrimidine antifolate 1 with activity against early- and late-stage pancreatic tumors with limited tumor selectivity. Structure-based design with our human serine hydroxymethyl transferase (SHMT) 2 and glycinamide ribonucleotide formyltransferase (GARFTase) structures, and published X-ray crystal structures of 5-aminoimidazole-4-carboxamide ribonucleotide formyltransferase/inosine monophosphate cyclohydrolase (ATIC), SHMT1, and folate receptor (FR) α and β afforded 11 analogues. Multitargeted inhibition and selective tumor transport were designed by providing promiscuous conformational flexibility in the molecules. Metabolite rescue identified mitochondrial C1 metabolism along with de novo purine biosynthesis as the targeted pathways. We identified analogues with tumor-selective transport via FRs and increased SHMT2, SHMT1, and GARFTase inhibition (28-, 21-, and 11-fold, respectively) compared to 1. These multitargeted agents represent an exciting new structural motif for targeted cancer therapy with substantial advantages of selectivity and potency over clinically used antifolates.