First-in-class multifunctional TYMS nonclassical antifolate inhibitor with potent in vivo activity that prolongs survival

Abstract

Although thymidylate synthase (TYMS) inhibitors have served as components of chemotherapy regimens, the currently available inhibitors induce TYMS overexpression or alter folate transport/metabolism feedback pathways that tumor cells exploit for drug resistance, limiting overall benefit. Here we report a small molecule TYMS inhibitor that i) exhibited enhanced antitumor activity as compared with current fluoropyrimidines and antifolates without inducing TYMS overexpression, ii) is structurally distinct from classical antifolates, iii) extended survival in both pancreatic xenograft tumor models and an hTS/Ink4a/Arf null genetically engineered mouse tumor model, and iv) is well tolerated with equal efficacy using either intraperitoneal or oral administration. Mechanistically, we verify the compound is a multifunctional nonclassical antifolate, and using a series of analogs, we identify structural features allowing direct TYMS inhibition while maintaining the ability to inhibit dihydrofolate reductase. Collectively, this work identifies nonclassical antifolate inhibitors that optimize inhibition of thymidylate biosynthesis with a favorable safety profile, highlighting the potential for enhanced cancer therapy.

Keywords: Cancer; Oncogenes; Therapeutics.

Figure 1. Compound 19-S inhibits TYMS catalytic activity, shows cytotoxicity in vitro, and does not increase TYMS levels.

(A) Diagram illustrating the TYMS (TS) tritium assay and conversion of tritiated dUMP to dTMP, generating tritiated water for quantification of TS activity. (B) TS tritium assay showing TS activity for control reactions and reactions performed in the presence of 5-fluorouridine (FUrd), 5-fluoro-2′-deoxyuridine (FdUrd), and compound 19-S. Drug concentrations were 250 μM using 10 μg/mL bacterially expressed TS protein. Mean ± SD of 4 data points from 2 independent experiments shown. The DMSO control represents 10 data points. Compound 19-S treatment represents 6 data points. (C) Viability assays comparing known TS inhibitor 5-FU with compound 19-S in the indicated cell lines following a 72-hour drug incubation. Data are expressed as mean ± SD of 2 independent experiments; n = 4 to 5 technical replicates. (D) Immunoblot analysis showing TS overexpression following 5-FU treatment (TS^5FU) and stable TS expression levels following compound 19-S treatment. PANC-1 cells were treated for 72 hours with the indicated 5-FU and 19-S concentrations. Experiment was repeated independently twice with similar results.

Figure 2. Compound 19-S inhibits tumor growth and progression.

(A) Experimental timeline for the subcutaneous (SQ) Luc-PANC-1 cell line–derived tumor model treated with compound 19-S or vehicle control. Compound 19-S (25 mg/kg) or vehicle control treatment cycles were administered by IP injection. For each cycle animals were treated once a day for 5 continuous days, then allowed 2 days’ rest when no treatments were administered. Tumor-bearing mice received a total of 4 treatment cycles before endpoint (n = 4 per cohort). (B) Effect of compound 19-S and vehicle control administered by IP injection on body weight as described in A. (C) Bioluminescence imaging of Luc-PANC-1–derived tumors and quantification of bioluminescence photon flux over time for animals treated with compound 19-S or vehicle control (**P = 0.0023). (D) Luc-PANC-1 tumor volumes for animals treated with compound 19-S or vehicle control (***P < 0.0001). (E) Final excised Luc-PANC-1 tumor weight for animals treated with compound 19-S or vehicle control (**P = 0.0029). (F) Timeline indicating the IP injection of Luc-CM cells to generate a disseminated tumor model and compound 19-S (25 mg/kg) or vehicle control treatment cycles administered by IP injection. For each cycle animals were treated once a day for 5 continuous days, then allowed 2 days’ rest when no treatments were administered. Tumor-bearing animals received 3 treatment cycles before endpoint (n = 8 per cohort). (G) Effect of compound 19-S and vehicle control administered by IP injection on body weight for Luc-CM tumor–bearing NSG mice. (H) Bioluminescence imaging of Luc-CM tumor–bearing NSG mice and quantification of bioluminescence photon flux from the abdominal region of animals treated with compound 19-S (n = 8) or vehicle control (n = 8). Data are expressed as mean ± SEM, ***P < 0.0001. Statistical analysis in C, D, and H was performed using 2-way ANOVA; in E, unpaired 2-tailed Student’s t test with Welch’s correction was used.

Figure 3. Compound 19-S prolongs survival in a pancreatic xenograft tumor model.

(A) Experimental timeline indicating the subcutaneous Luc-PANC-1 cell line–derived tumor model treated with compound 19-S or vehicle control. Timeline indicates the subcutaneous injection of Luc-PANC-1 cells in NSG mice to generate tumors and compound 19-S (10 mg/kg or 25 mg/kg) or vehicle control treatment cycles administered by oral gavage (per os; PO). For each cycle animals were treated once a day for 5 continuous days, then allowed 2 days’ rest when no treatments were administered. Animals received treatment until survival endpoint. (B) Effect of compound 19-S and vehicle control administered PO on body weight for Luc-PANC-1 tumor–bearing NSG mice. Data are presented as mean body weight of 5 animals per group ± SEM. (C) Kaplan-Meier survival analysis for Luc-PANC-1–injected NSG mice treated with compound 19-S (10 mg/kg or 25 mg/kg, n = 5 per group) or vehicle control (n = 5); **P = 0.003. Log-rank (Mantel-Cox) test was used to calculate P values. (D and E) Bioluminescence imaging of Luc-PANC-1–derived tumors and quantification of bioluminescence photon flux over time for animals treated with compound 19-S (10 mg/kg or 25 mg/kg) or vehicle control. (F) Luc-PANC-1 tumor volumes for animals treated with compound 19-S (10 mg/kg or 25 mg/kg) or vehicle control. In E and F, data are presented as mean total flux or tumor volume, respectively, of 5 animals per group ± SEM; **P ≤ 0.01 by 2-tailed Mann-Whitney t test.

Figure 4. Compound 19-S prolongs survival in an hTS/Ink4a/Arf ^–/– GEMM.

(A) Generation of hTS/Ink4a/Arf^–/– mice by crossing hTS transgenic mice with Ink4a/Arf^–/– mice. Locations of forward and reverse primers for the detection of hTS transgene and Ink4a/Arf locus are shown by arrows. (B) Experimental timeline for hTS/Ink4a/Arf^–/– GEMM survival experiments. Treatment started when animals were 3 months of age, and a total of 4 treatment cycles were administered. For each treatment cycle, animals were administered compound 19-S (10 mg/kg) or vehicle control by IP injection twice weekly for 3 weeks and then allowed 1-week rest with no treatment. Animals were then monitored until survival endpoint. (C) Kaplan-Meier survival analysis for hTS/Ink4a/Arf ^–/– animals treated with compound 19-S (n = 26) or vehicle control (n = 25). ***P < 0.0001 was calculated by log-rank (Mantel-Cox) test. Tx, treatment. (D) Diagrams indicating the surviving and deceased fractions for the compound 19-S treatment group and vehicle control group at the end of treatment cycle 4, when treatment was discontinued for all animals.

Figure 5. Synthesis of compound 19-S analogs.

Reagents and conditions: synthesis of 19-S2: pyrimethamine, HNO₃, H₂SO₄, 0°C to 50°C, 80 minutes, 99% yield. Synthesis of 19-S, 19-S1, 19-S3 through 19-S9, and 19-S11 through 19-S14: 19-S2, amine (neat), 150°C, sealed tube, 6 hours, 22–86% yield. Synthesis of 19-S10: 19-S9, TFA, CH₂Cl₂, room temperature, 80% yield. TFA, trifluoroacetic acid.

Figure 6. Select compound 19-S series analogs show dual TS and DHFR inhibition.

(A) Tritium-based TS activity assay screen of compound 19-S series analogs. Compounds were initially screened at 250 μM to determine analogs showing TS inhibition. Mean ± SD of 4 data points from 2 independent experiments shown. DMSO, TS + 5,10-mTHF, and 5,10-mTHF controls represent 8 data points. P values are calculated using 2-tailed unpaired t test, ****P ≤ 0.0001. (B) Absorbance-based DHFR activity assay screen of compound 19-S series analogs; all compounds were initially screened at 1 μM concentration. Data are expressed as mean ± SD of n = 7 from 2 independent experiments. (C) Comparison showing TS activity utilizing the tritium-based activity assay in the presence of pemetrexed (PEM), methotrexate (MTX), and compounds 19-S, 19-S5, and 19-S7 at the indicated concentrations. Data are expressed as mean ± SD of n = 4 from 2 independent experiments. (D) Comparison showing DHFR activity utilizing the absorbance-based activity assay in the presence of PEM, MTX, and compounds 19-S, 19-S5, and 19-S7 at the indicated concentrations. Data are expressed as mean ± SD of n = 4. (E) MIA PaCa-2 cell line viability assays following 72-hour drug treatment with compound 19-S and 19-S series analogs 19-S5 and 19-S7 and known control antifolate inhibitors PEM and MTX. Data are expressed as mean ± SD (triplicates).

Figure 7. Mechanism of TS inhibition reveals nonclassical antifolate inhibitors.

(A) Illustration outlining the drug displacement from TS with increasing substrate concentrations to highlight how increased molar ratios of the competing substrate will displace the prebound drug and shift the equilibrium to the enzyme/substrate complex. resulting in an increased conversion of tritiated dUMP. (B–E) Tritium assay for TS activity with increasing 5,10-mTHF or dUMP ratios for 19-S (19-S) (B), 19-S5 (19-S5) (C), 19-S7 (19-S7) (D), and control classical antifolate PEM (E). TS was preincubated with each compound using the indicated concentration; for each compound the concentration required for a 50% reduction in dUMP conversion after a 30-minute reaction was utilized to allow changes in dUMP conversion to be observed. Increasing TS activity with increasing 5,10-mTHF ratios was observed for all compounds, including the control antifolate PEM, while increasing dUMP ratios had no significant effect on TS activity, indicating compounds 19-S, 19-S5, and 19-S7 act as nonclassical antifolate inhibitors by competing only with the 5,10-mTHF substrate. Data are expressed as mean ± SD of n = 4 from 2 independent experiments.

Figure 8. Proposed binding mode for compound 19-S.

(A–D) Compound 19-S at canonical binding sites of TS and DHFR, respectively. A and B show the protein backbone as gray and purple ribbons for TS and DHFR, respectively. Cofactors, key residues, and compound 19-S are depicted as sticks and colored by element with carbons, oxygens, nitrogen atoms, and phosphorous in orange, red, blue, and brown, respectively (except compound 19-S with carbons in cyan). C and D show a 2D diagram of the protein-ligand interactions between compound 19-S and TS and DHFR, respectively. Key residues and 19-S are colored by elements with carbons, oxygens, and nitrogen atoms in black, red, and blue, respectively. Polar and aromatic interactions are presented as black and green dotted dashed lines, respectively. Nonpolar interactions are presented as a continuous green line surrounding the ligand functional groups.

Figure 9. Compounds 19-S and 19-S7 prevent tumor progression without signs of toxicity following oral delivery.

(A) Timeline indicating the subcutaneous injection of Luc-PANC-1 cells to generate subcutaneous tumors and the treatment cycles for compound 19-S (25 mg/kg), compound 19-S7 (25 mg/kg), or vehicle control administered PO. For each cycle animals were treated once a day for 5 continuous days, then allowed 2 days’ rest when no treatments were administered. NSG mice received 4 treatment cycles until endpoint. (B) Effect of compound 19-S and 19-S7 treatment with matched vehicle control treatment on body weight for Luc-PANC-1 tumor–bearing NSG mice. Data are expressed as mean ± SEM of compound 19-S (n = 7), 19-S7 (n = 6), and controls (n = 7 and n = 6, respectively). (C) Bioluminescence imaging of Luc-PANC-1–derived tumors and quantification of bioluminescence photon flux over time for animals treated PO with compound 19-S or vehicle control (***P = 0.0002). (D) Luc-PANC-1 tumor volumes for animals treated with compound 19-S or vehicle control administered PO (***P < 0.0001). (E) Final excised Luc-PANC-1 tumor weight for animals treated PO with compound 19-S or vehicle control (***P = 0.0010). (F) Bioluminescence imaging of Luc-PANC-1 derived tumors and quantification of bioluminescence photon flux over time for animals treated PO with compound 19-S7 or vehicle control (*P = 0.0192). (G) Luc-PANC-1 tumor volumes for animals treated PO with compound 19-S7 or vehicle control (***P < 0.0001). (H) Final excised Luc-PANC-1 tumor weight for animals treated PO with compound 19-S7 or vehicle control, ***P = 0.0010. Data are expressed as mean ± SEM of n = 7 for 19-S and n = 6 for 19-S7 treatment. Statistical analysis in C, D, F, and G was performed using 2-way ANOVA; in E and H unpaired t test with Welch’s correction was used.