Repurposing the Antidepressant Sertraline as SHMT Inhibitor to Suppress Serine/Glycine Synthesis-Addicted Breast Tumor Growth

Abstract

Metabolic rewiring is a hallmark of cancer that supports tumor growth, survival, and chemotherapy resistance. Although normal cells often rely on extracellular serine and glycine supply, a significant subset of cancers becomes addicted to intracellular serine/glycine synthesis, offering an attractive drug target. Previously developed inhibitors of serine/glycine synthesis enzymes did not reach clinical trials due to unfavorable pharmacokinetic profiles, implying that further efforts to identify clinically applicable drugs targeting this pathway are required. In this study, we aimed to develop therapies that can rapidly enter the clinical practice by focusing on drug repurposing, as their safety and cost-effectiveness have been optimized before. Using a yeast model system, we repurposed two compounds, sertraline and thimerosal, for their selective toxicity against serine/glycine synthesis-addicted breast cancer and T-cell acute lymphoblastic leukemia cell lines. Isotope tracer metabolomics, computational docking, enzymatic assays, and drug-target interaction studies revealed that sertraline and thimerosal inhibit serine/glycine synthesis enzymes serine hydroxymethyltransferase and phosphoglycerate dehydrogenase, respectively. In addition, we demonstrated that sertraline's antiproliferative activity was further aggravated by mitochondrial inhibitors, such as the antimalarial artemether, by causing G₁-S cell-cycle arrest. Most notably, this combination also resulted in serine-selective antitumor activity in breast cancer mouse xenografts. Collectively, this study provides molecular insights into the repurposed mode-of-action of the antidepressant sertraline and allows to delineate a hitherto unidentified group of cancers being particularly sensitive to treatment with sertraline. Furthermore, we highlight the simultaneous inhibition of serine/glycine synthesis and mitochondrial metabolism as a novel treatment strategy for serine/glycine synthesis-addicted cancers.

Figure 1. Schematic overview of study design.

In terms of serine/glycine metabolism, breast cancers (BRCA) can largely be divided into serine/glycine uptake or synthesis addicted. Using a yeast model system that upregulates serine/glycine synthesis, we selected repurposed compounds that target serine/glycine synthesis in the breast cancer context. After thorough in vitro validation, the most promising, and clinically used, repurposed compound was selected for the rational design of a novel combination therapy for serine/glycine synthesis addicted breast cancer. Human enzymes involved in serine/glycine synthesis are indicated in blue. PHGDH: phosphoglycerate dehydrogenase; PSAT1: phosphoserine aminotransferase; PSPH: phosphoserine phosphatase; SHMT1/2: cytosolic/mitochondrial serine hydroxymethyltransferase; TCA: tricarboxylic acid.

Figure 2. A subset of “re-sensitizing” agents impairs proliferation of serine/glycine synthesis addicted breast cancer cell lines.

Proliferation during 96 hours, as determined by real-time monitoring of cell confluence (%), of MDA-MB-231 (upper) and MDA-MB-468 (lower) cells upon treatment with indicated concentrations of sertraline, thimerosal, benzalkonium chloride and bupropion (left to right). One representative result of three biological replicates, containing each at least three technical replicates, is shown (mean ± SD).

Figure 3. Sertraline and thimerosal target serine/glycine synthesis.

(A) Survival, by measuring cell viability using flow cytometry, of Ba/F3 cells, that express either RPL10 WT or RPL10 R98S, upon treatment with 7.3 μM sertraline (left) or 1 μM thimerosal (right) for 48 hours. Values are presented relative to the control treatment (n = 3 individual CRISPR/Cas9 clones with at least two technical replicates in each experiment, Student’s t-test). (B) Proliferation, as determined by real-time monitoring of cell confluence (%), of MDA-MB-468 cells cultured in DMEM with (upper) or without (lower) serine (400 μM) and treated with indicated concentrations of sertraline (left) or thimerosal (right). One representative result of three biological replicates, containing each at least three technical replicates, is shown. (C) Relative abundance of intracellular serine and glycine in MDA-MB-468 cells treated with indicated concentrations of sertraline (upper) and thimerosal (lower) for 72 and 24 hours, respectively (n = 3, One-way ANOVA, Dunnett’s multiple comparisons test). (D) Schematic representation of carbon incorporation from ¹³C₆-glucose into serine and glycine. Glucose-derived serine and glycine shows mass shifts of 3 and 2 units (M+3 serine and M+2 glycine), respectively. Cellular uptake of serine and glycine from the cell culture medium will result in unlabeled (M+0) serine and glycine, whereas interconversion between serine and glycine, catalyzed by SHMT1/2, will result in partially labeled serine (M+1 and M+2) and glycine (M+1). (E) Serine and glycine mass distribution showing the fractional glucose contribution of each mass upon treatment of MDA-MB-468 cells with indicated concentrations of sertraline (upper) and thimerosal (lower) (n = 3, One-way ANOVA, Dunnett’s multiple comparisons). In (A-C and E) data are presented as mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001.

Figure 4. Thimerosal reduces PHGDH activity, while sertraline inhibits SHMT1/2.

(A) PHGDH enzymatic in vitro assay, measuring PHGDH activity upon addition of indicated concentrations of sertraline (left) and thimerosal (right). Values are presented relative to the control (n = 3). (B) SHMT1 (left) and SHMT2 (right) in complex with sertraline (grey), with a magnified view of the binding pocket showing the interactions formed by sertraline. H-bonds formed by sertraline are presented as yellow dashes. The known SHMT inhibitor SHIN1, with a pyrazolopyran scaffold, is shown in magenta. (C) Schematic overview of isotopic tracing with [2,3,3-²H]-serine, showing ²H incorporation in downstream metabolites glycine and thymidine (dTTP). Cells taking up fully deuterated (M+3) serine use this to synthesize glycine with one deuterium label (M+1). While cytosolic methylene-THF production, by SHMT1, results in double ²H-labeled (M+2) dTTP (red dots), mitochondrial SHMT2 will generate single ²H-labeled (M+1) dTTP (blue dots). (D) Serine mass distribution showing the labeling fraction of each mass upon treatment of MDA-MB-468 cells with control (DMSO) or sertraline (5 μM) for 48 hours (n = 3, Multiple t-test). (E) Deuterium M+1 labeled dTTP fraction in MDA-MB-468 cells treated with control (DMSO) or sertraline (5 μM) for 48 hours. Values are presented relative to the control (n = 3, Student’s t-test). (F) Fluorescence distribution curves (left) and dose response curve (right) for SHMT2 incubated with different concentrations of sertraline (0.36 μM - 730 μM). K_d = 13.1 μM. Error bars represent the standard deviation of the measurements of two independent repeats. (G) Melting temperature (Tm) curves demonstrating sertraline-induced destabilization of SHMT1. One representative result of two biological replicates is shown. (H) Serine and glycine uptake (-) and secretion (+) rates (AU/cell/h) of MDA-MB-468 cells treated with indicated concentrations of sertraline for 72 h (n = 3, One-way ANOVA, Dunnett’s multiple comparisons). In (A, D-F and H) data are presented as mean ± SD. *p < 0.05, **p < 0.01.

Figure 5. Sertraline has clinical potential, especially in combination with mitochondrial inhibitors.

(A) Proliferation during 96 hours, as determined by real-time monitoring of cell confluence (%), of MDA-MB-231 (upper) and MDA-MB-468 (lower) cells upon treatment with sertraline (5 μM) in combination with rotenone (50 nM), antimycin A (50 nM) or artemether (80 μM). One representative result of three biological replicates, containing each at least three technical replicates, is shown. (B) Histograms showing PI cell cycle analysis (left) and BrdU incorporation (right) of MDA-MB-468 cells treated with DMSO, sertraline (5 μM) and/or artemether (80 μM) for 24 hours. One representative result of three biological replicates is shown. (C) Quantification of (B) pooling all three biological replicates (n = 3, Two-way ANOVA, Dunnett’s multiple comparisons test). (D) Tumor weight (g) of MDA-MB-231 (left flank) and MDA-MB-468 (right flank) mouse xenografts after treatment with DMSO, sertraline (2.5 mg/kg), artemether (40 mg/kg) or a combination of both compounds for 4 weeks (n ≥ 3, Kruskal-Wallis test, followed by Mann-Whitney U test). (E) Schematic overview of the mode-of-action of sertraline and thimerosal. In (A, C and D) data are presented as mean ± SD. *p < 0.05, **p < 0.01, ****p < 0.0001.