Impairments in SHMT2 expression or cellular folate availability reduce oxidative phosphorylation and pyruvate kinase activity

Abstract

Background: Serine hydroxymethyltransferase 2 (SHMT2) catalyzes the reversible conversion of tetrahydrofolate (THF) and serine-producing THF-conjugated one-carbon units and glycine in the mitochondria. Biallelic SHMT2 variants were identified in humans and suggested to alter the protein's active site, potentially disrupting enzymatic function. SHMT2 expression has also been shown to decrease with aging in human fibroblasts. Immortalized cell models of total SHMT2 loss or folate deficiency exhibit decreased oxidative capacity and impaired mitochondrial complex I assembly and protein levels, suggesting folate-mediated one-carbon metabolism (FOCM) and the oxidative phosphorylation system are functionally coordinated. This study examined the role of SHMT2 and folate availability in regulating mitochondrial function, energy metabolism, and cellular proliferative capacity in both heterozygous and homozygous cell models of reduced SHMT2 expression. In this study, primary mouse embryonic fibroblasts (MEF) were isolated from a C57Bl/6J dam crossed with a heterozygous Shmt2^+/- male to generate Shmt2^+/+ (wild-type) or Shmt2^+/- (HET) MEF cells. In addition, haploid chronic myeloid leukemia cells (HAP1, wild-type) or HAP1 cells lacking SHMT2 expression (ΔSHMT2) were cultured for 4 doublings in either low-folate or folate-sufficient culture media. Cells were examined for proliferation, total folate levels, mtDNA content, protein levels of pyruvate kinase and PGC1α, pyruvate kinase enzyme activity, mitochondrial membrane potential, and mitochondrial function.

Results: Homozygous loss of SHMT2 in HAP1 cells impaired cellular folate accumulation and altered mitochondrial DNA content, formate production, membrane potential, and basal respiration. Formate rescued proliferation in HAP1, but not ΔSHMT2, cells cultured in low-folate medium. Pyruvate kinase activity and protein levels were impaired in ΔSHMT2 cells and in MEF cells exposed to low-folate medium. Mitochondrial biogenesis protein levels were elevated in Shmt2^+/- MEF cells, while mitochondrial mass was increased in both homozygous and heterozygous models of SHMT2 loss.

Conclusions: The results from this study indicate disrupted mitochondrial FOCM impairs mitochondrial folate accumulation and respiration, mitochondrial formate production, glycolytic activity, and cellular proliferation. These changes persist even after a potentially compensatory increase in mitochondrial biogenesis as a result of decreased SHMT2 levels.

Keywords: Energy metabolism; Folate; One-carbon metabolism; Oxygen consumption rate; Pyruvate kinase; SHMT2.

Fig. 1

Total folates, cellular uptake and particianing of folate, mtDNA content, mitochondrial membrane potential, and basal respirationin HAP1 and ΔSHMT2 cells. Homozygous loss of SHMT2 and low-folate medium decrease folate accumulation and basal respiration and increase mitochondrial membrane potential in HAP1 cells. A SHMT2 protein levels, B total folate levels, C cellular uptake and partitioning of folate, D mtDNA content, E mitochondrial membrane potential, and F oxygen consumption rate in HAP1 cells and ΔSHMT2 cells. SHMT2 protein levels were normalized to GAPDH, and densitometry was performed using ImageJ. Two-way ANOVA with Tukey’s post hoc analysis was used to determine media by genotype interaction and main effects of media and genotype with a statistical significance at p < 0.05. Levels not connected by the same letter are significantly different. Data represent means ± SD values, n = 2–6 per group. GAPDH, glyceraldegyde-3 phosphate dehydrogenase; SHMT2, serine hydroxymethyltransferase 2

Fig. 2

Cellular proliferation rates in HAP1 and ΔSHMT2 cells. Formate rescues cell proliferation rate in HAP1 cells cultured in low-folate medium but not in ΔSHMT2 cells. Cell proliferation rates of ΔSHMT2 cells were compared with HAP1 cells by co-staining cells with Hoechst 33342 (to identify all cells) and propidium iodide (to identify dead cells). Fold change of each group was calculated by dividing by day 0 cell number. Data represent means ± SD values. Values represent n = 6 replicates of cell lines cultured in medium containing either 25 nM (6S)5-formyl-THF or 0 nM (6S)5-formyl-THF. A Cell proliferation rate and cell proliferation rate in the presence of 2 mM formate, B relative days 1–4 quantitation of cell proliferation rate, and C relative days 1–4 quantitation cell proliferation rate in the presence of 2-mM formate. Linear mixed-effects models with main effects of media, genotype, and time (with time as a continuous variable) and 2- and 3-way interactions were used to determine cell proliferation with a statistical significance at p < 0.05. Two-way ANOVA with Tukey’s post hoc analysis was used to determine media by genotype interaction and main effects of media, and genotype with a statistical significance at p < 0.05 were used to analyze individual day proliferation. Levels not connected by the same letter are significantly different

Fig. 3

Contribution of serine one-carbon units for de novo dTMP synthesis. _L-[2,3,4-²H₃]-serine contribution to dTMP synthesis from formate was reduced with homozygous loss of SHMT2 and in HAP1 cells cultured in low-folate medium compared to HAP1 cells culture in folate-sufficient medium. One-way ANOVA with Tukey’s post hoc analysis was used with a statistical significance of p < 0.05. Data represents means ± SD values, n = 2–3 per group. _L-[2,3,4-²H₃]-serine incorporation into thymidine is conveyed by the ratio of carbons containing one deuterium atom/total number of carbons containing one and two deuterium atoms

Fig. 4

Protein levels and pyruvate kinase activity in HAP1 and ΔSHMT2 cells. Homozygous loss of SHMT2 reduces protein levels and activity of pyruvate kinase in HAP1 cells. A PKM1 and PKM2 protein levels and B pyruvate kinase activity in HAP1 cells and ΔSHMT2 cells. PKM1 and PKM2 protein levels were normalized to GAPDH, and densitometry was performed using ImageJ. Two-way ANOVA with Tukey’s post hoc analysis was used to determine media by genotype interaction and main effects of media and genotype with a statistical significance at p < 0.05. Levels not connected by the same letter are significantly different. Data represent means ± SD values, n = 2–5 per group. GAPDH, glyceraldegyde-3 phosphate dehydrogenase; PKM1 and PKM2, pyruvate kinase M1 and M2 isoforms

Fig. 5

Protein levels and pyruvate kinase activity inShmt2^+/+ and Shmt2^+/− MEF cells. Low-folate medium decreases protein levels, and activity of pyruvate kinase in MEF cells and reduced Shmt2 expression increases PGC1α protein levels. A PKM1, PKM2, and PGC1α protein levels and B pyruvate kinase activity in Shmt2^+/+ and Shmt2^+/− MEF cells. PKM1, PKM2, and PGC1α protein levels were normalized to GAPDH, and densitometry was performed using ImageJ. Two-way ANOVA with Tukey’s post hoc analysis was used to determine media by genotype interaction and main effects of media and genotype with a statistical significance at p < 0.05. Levels not connected by the same letter are significantly different. Data represent means ± SD values, n = 2–4 per group with 2 embryo cells lines represented in each group. GAPDH, glyceraldegyde-3 phosphate dehydrogenase; PGC1α, PPARγ coactivator-1α; PKM1 and PKM2, pyruvate kinase M1 and M2 isoforms

Fig. 6

Cell type ATP production and extracellular acidification rates. Decreased SHMT2 and exposure to low-folate medium impair ATP production and ECAR. ATP production and extracellular acidification rates in A HAP1 cells and ΔSHMT2 cells and B Shmt2^+/+ and Shmt2^+/− MEF cells. ATP production and extracellular acidification rates were normalized to total cell count. Two-way ANOVA with Tukey’s post hoc analysis was used to determine media by genotype interaction and main effects of media and genotype with a statistical significance at p < 0.05. Levels not connected by the same letter are significantly different. Data represent means ± SD values, n = 4 per group with 2 embryo cells lines represented in each group

Fig. 7

Cell type responses in mitochondrial mass. Decreased SHMT2 leads to increased mitochondrial mass. Citrate synthase activity in A HAP1 cells and ΔSHMT2 cells and B Shmt2^+/+ and Shmt2^+/− MEF cells. Citrate synthase activity was normalized to total protein. Two-way ANOVA with Tukey’s post hoc analysis was used to determine media by genotype interaction and main effects of media and genotype with a statistical significance at p < 0.05. Levels not connected by the same letter are significantly different. Data represent means ± SD values, n = 4 per group with 2 embryo cells lines represented in each group. CS, citrate synthase

Fig. 8

Cell type responses in NAD/NADH ratio. NAD/NADH ratio is impaired with homozygous SHMT2 loss but not heterozygous loss. NAD/NADH ratio in A HAP1 cells and ΔSHMT2 cells and B Shmt2^+/+ and Shmt2^+/− MEF cells. NAD/NADH ratio was normalized to total cell count. Student’s t-test was used to determine genotype effects with a statistical significance (*) at p < 0.05. Data represent means ± SD values, n = 4–8 per group with 2 embryo cells lines represented in each group for the MEF cell analyses