Mitochondrial One-Carbon Pathway Supports Cytosolic Folate Integrity in Cancer Cells

Abstract

Mammalian folate metabolism is comprised of cytosolic and mitochondrial pathways with nearly identical core reactions, yet the functional advantages of such an organization are not well understood. Using genome-editing and biochemical approaches, we find that ablating folate metabolism in the mitochondria of mammalian cell lines results in folate degradation in the cytosol. Mechanistically, we show that QDPR, an enzyme in tetrahydrobiopterin metabolism, moonlights to repair oxidative damage to tetrahydrofolate (THF). This repair capacity is overwhelmed when cytosolic THF hyperaccumulates in the absence of mitochondrially produced formate, leading to THF degradation. Unexpectedly, we also find that the classic antifolate methotrexate, by inhibiting its well-known target DHFR, causes even more extensive folate degradation in nearly all tested cancer cell lines. These findings shed light on design features of folate metabolism, provide a biochemical basis for clinically observed folate deficiency in QDPR-deficient patients, and reveal a hitherto unknown and unexplored cellular effect of methotrexate.

Keywords: folate; formate; metabolism; methotrexate; mitochondria; one-carbon.

Figure 1. Formation of an Unusual Folate Metabolite in Mitochondrial-1C-Pathway-Deficient Cells

(A) Compartmentation of mammalian 1C metabolism. SLC25A32 transports only the monoglutamate form of folates. In this intercompartmental cycle, 1C tends to flow clockwise, with serine oxidation in mitochondria and formate reduction in the cytosol. (B) Folate-centered view of the mitochondrial 1C pathway. (C) Western blot analysis of MDA-MB-468 cells stably expressing non-targeting or SHMT2-targeting short hairpin RNAs (shRNAs). (D) Western blot analysis of MDA-MB-468 parental and MTHFD1L-knockout cells with or without ectopic expression of MTHFD1L. (E) Western blot analysis of MDA-MB-468 parental and MTHFD1-knockout cells with or without ectopic expression of MTHFD1. The D125A mutation ablates the dehydrogenase and cyclohydrolase activity of MTHFD1. (F–N) HPLC analysis of folates from isogenic MDA-MB-468 lines. ‘‘+ MTHFD1L’’ denotes ectopic MTHFD1L expression; ‘‘+ formate’’ denotes supplementing the labeling medium with 2 mM sodium formate; ‘‘+ hypoxanthine’’ denotes supplementing the labeling medium with 100 μM hypoxanthine, a necessity because the MTHFD1-knockout cells are purine auxotrophic. ‘‘+ MTHFD1’’ denotes ectopic MTHFD1 expression. See also Figure S1.

Figure 2. Formation of the Same Unusual Folate Metabolite upon Methotrexate Treatment

(A) HPLC analysis of folates from MDA-MB-468 cells, untreated or treated with 1 μM methotrexate for indicated times. (B) Western blot analysis of 293T parental, DHFR-knockout, DHFR-DHFRL1-double-knockout, and DHFR-TYMS-double-knockout cells. (C) HPLC analysis of folates from 293T parental, DHFR-knockout, DHFR-DHFRL1-double-knockout cells, untreated or treated with 1 μM methotrexate for 24 hr. 100 μM hypoxanthine and 16 μM thymidine (HT) were included in the medium to support cell growth. (D) HPLC analysis of folates from 293T DHFR-TYMS-double-knockout cells. (E) HPLC analysis of folates from MDA-MB-468 TYMS-knockout cells, untreated or treated with 1 μM methotrexate for 6 hr. 16 μM thymidine was included in the medium to support cell growth. See also Figure S2.

Figure 3. Identification of the Unusual Folate Metabolite

(A) Schematic of the modified ‘‘pulse-chase’’ experimental design. (B) HPLC quantification of folates at the indicated chase times. Data points are normalized to summed peak area of all known folates at 0.5 hr. Error bars depict SD. (C) Schematic of pAcABG_n hydrolysis facilitated by GGH, or acid plus heat. pABA, p-aminobenzoic acid; pABG, p-aminobenzoyl-glutamate. (D) ³H-labeled, GGH-treated, HPLC-purified un-pAcABG usual metabolite was mixed with unlabeled pAcABG standard, untreated or treated with 1 N HCl at 99°C for 5 min or 1 hr, and analyzed by HPLC. The radioactive unusual metabolite (⁓5 Ci/mmol) displayed negligible UV absorption. See also Figure S3.

Figure 4. Roles of Nonenzymatic Oxidation and NAT1 in Folate Catabolism

(A) HPLC analysis of folates from MDA-MB-468 non-targeting control and SHMT2-knockdown cells, untreated or treated with 1 mM H₂O₂ for 5 hr. The asterisk denotes a 5-CH₃-THF-derived metabolite, likely to be a 5-methyl-dihydro-pyrazino-s-trazine compound (Gregory, 1989). (B) Western blot analysis of MDA-MB-468 parental and NAT1-knockout cells generated using two independent CRISPR guides. (C) HPLC analysis of folates from the NAT1-knockout cells, untreated (top), treated with 1 μM methotrexate for 6 hr (middle), or treated with 1 μM methotrexate for 6 hr and then after cell lysis, further treated with recombinant DHFR and NADPH (bottom). (D) Schematic of THF catabolism. ‘‘[O]’’ indicates the oxidative and presumably nonenzymatic nature of the cleavage step. See also Figure S4.

Figure 5. QDPR Repairs Oxidative Damage to Unsubstituted THF in Cells

(A) Distinct chemical degradation and enzymatic repair pathways for THF and 10-CHO-THF. (B) QDPR prevents 6(RS)-THF from decomposition at 37° C in vitro in an NADH-dependent manner. (C) Unlike QDPR, DHFR does not prevent THF from decomposition at 37° C in vitro. (D–L) HPLC analysis of folates from the indicated cells. 100 μM hypoxanthine was included in the medium to support the growth of the MTHFD1-knockout cells, which exhibit purine auxotrophy. (M) Schematic of the serine-deprivation experiment (in [N]–[P]). (N–P) Differential kinetics of THF decomposition into pAcABG in MDA-MB-468 parental, QDPR-knockout, and QDPR-overexpressing cells. (Q) Differential degrees of pAcABG accumulation in MDA-MB-468 parental, QDPR-knockout, and QDPR-overexpressing cells, untreated or treated with 1 mM H₂O₂ for 3 hr. Error bars depict SD. See also Figure S5.

Figure 6. QDPR Genetically Interacts with DHFR, and ALDH1L1 Expression Alters the Folate Response to Methotrexate

(A and B) Time course of folate interconversion and decomposition in MDA-MB-468 parental (A) and QDPR-overexpressing (B) cells upon 1 μM methotrexate treatment. (C) Western blot analysis for the cellular thermal shift assay on MDA-MB-468 cells treated with 1 μM methotrexate for 30 min. (D–L) HPLC analysis of folates from the indicated cells. (M) ALDH1L1 protects 6(S)-THF from decomposition at 37°C in vitro. (N) ALDH1L1 converts DHF into folic acid in aerobic buffers at 37°C. (O and P) Time course of folate interconversion and decomposition in MDA-MB-468 parental and ALDH1L1-overexpressing cells upon 1 μM methotrexate treatment. (Q and R) HPLC analysis of folates from the bladder cancer cell line RT4, untreated or treated with 1 μM methotrexate for 4 hr. See also Figure S6.

Figure 7. Glycine Auxotrophy Is Rescued by Overexpression of ALDH1L Enzymes

(A–I) Growth curves of MDA-MB-468 parental, MTHFD1L-knockout, and SLC25A32-knockout cells, with or without ALDH1L1 or ALDH1L2 overexpression, and with or without supplementation with 400 μM glycine or 100 μM hypoxanthine. Adding 16 μM thymidine on top of 100 μM hypoxanthine had no additional effects (data not shown). Error bars depict SD. (J) Western blot analysis of 293T parental, MTHFD1L-knockout, and MTHFD1L-ALDH1L2-double-knockout cells. The double-knockout clones (A1) and (B1) were derived from the MTHFD1L-knockout clone using two independent CRISPR guides against ALDH1L2. (K) Representative images of the cells in (J) plated in glycine-free medium. Scale bars: 400 μm. See also Figure S7.