Elucidating the mechanisms by which disulfiram protects against obesity and metabolic syndrome

Abstract

There is an unmet need and urgency to find safe and effective anti-obesity interventions. Our recent study in mice fed on obesogenic diet found that treatment with the alcohol aversive drug disulfiram reduced feeding efficiency and led to a decrease in body weight and an increase in energy expenditure. The intervention with disulfiram improved glucose tolerance and insulin sensitivity, and mitigated metabolic dysfunctions in various organs through poorly defined mechanisms. Here, integrated analysis of transcriptomic and proteomic data from mouse and rat livers unveiled comparable signatures in response to disulfiram, revealing pathways associated with lipid and energy metabolism, redox, and detoxification. In cell culture, disulfiram was found to be a potent activator of autophagy, the malfunctioning of which has negative consequences on metabolic regulation. Thus, repurposing disulfiram may represent a potent strategy to combat obesity.

Keywords: Metabolic syndrome; Obesity.

Conflict of interest statement

Competing interestsThe authors declare no competing interests.

© This is a U.S. government work and not under copyright protection in the U.S.; foreign copyright protection may apply 2020.

Figures

Fig. 1. Disulfiram significantly modifies the liver transcriptome profile in HFD-fed mice.

a Experimental design. Vehicle, black bar; low dose of DSF, blue bar; high dose of DSF, red bar. b Body weight of the six experimental groups of animals at the start and the conclusion of the study. n = 14–22 per group. c Principal component analysis (PCA) was performed on liver of mice fed a standard (SD) or high-fat diet (HFD) supplemented or not with low and high doses of DSF for 41 weeks. d Venn diagram of significantly upregulated (red font), downregulated (blue font), and reciprocally regulated (black font) gene transcripts. e Graphical representation of the 80 genes reciprocally regulated in the HFD-SD, HFDL-HFD, and HFDH-HFD (data not shown) pairwise comparisons. Cyp2b13, Cidec, Hk2, Cyp7b1, Igfbp2, and Hsd3b5 were among the top reciprocally regulated liver genes. Additional information is provided in Supplemental Table 3. f Venn diagram depicting the distribution of GO Terms with positive (red font) and negative (blue font) z-ratios derived from the HFD-SD, HFDL-HFD, and HFDH-HFD pairwise comparisons. The number of GO Terms in black represents z-ratios in opposite direction between the three pairwise comparisons. g A select group of canonical pathways enriched in genes significantly impacted in the HFD-SD and HFDL-HFD pairwise comparisons. h Validation of the microarray data by quantitative real-time PCR. n = 4. i Liver extracts were prepared from mice after 41 weeks of dietary intervention and then immunoblotted for IGFBP2 (left panel). Relative protein expression after data normalization using Ponceau S staining of the membrane is depicted in right panel. Data in f, g are shown as mean ± SEM. *P ≤ 0.05 compared with diet without DSF. Related to Supplementary Fig. 1 and Supplementary Tables 1–3 and 6.

Fig. 2. DSF reduces oxidative stress and acetylation markers in HFD-fed mice.

a Detection of acetylated and total forms of SOD2, SIRT1, and IL-1β proteins in liver homogenates. b Relative protein expression (acetylated/total SOD2 ratio and IL-1β) after data normalization using GAPDH as loading control. c Detection of 4-HNE-conjugated proteins in liver homogenates. Ponceau S staining of the membrane confirmed equal protein load (lower panel). d Densitometric analysis of 4-HNE signals after normalization with Ponceau S. Data are shown as mean ± SEM. *P ≤ 0.05 compared with diet without DSF; #P ≤ 0.05 compared with low DSF. Related to Supplementary Fig. 2.

Fig. 3. Multi-omics analysis of the effects of disulfiram in rats.

Proteomic analysis of liver proteins (a–g) and untargeted serum metabolomics (h–j) in rats exposed to disulfiram. a Heat map visualization of protein abundance in livers of rats fed laboratory SD (control) either supplemented or not with a low or high dose of DSF. Upregulation (red font), downregulation (blue font). n = 8 per group. b Volcano plots of liver protein abundance changes after diet supplementation with low or high dose of DSF were plotted with the y-axis showing the Benjamini–Hochberg corrected −log10 (P) and the x-axis showing the log2 fold change of protein abundance (DSF/control SD) calculated from the median LFQ intensity values. The blue and red symbols denote significant changes with low and high DSF, respectively, and the black symbols denote nonsignificant changes. Significance is defined as >40% median fold changes in either direction. c Two-way Venn diagram depicting the distribution of unique and common proteins whose expression was impacted by DSF (low or high dose) vs. control SD. Upregulation (red font), downregulation (blue font). Significance is defined as >40% median fold changes in either direction. d Abundance of a select group of proteins significantly impacted by DSF treatment. Normalized LFQ intensity values are represented in box and whisker plot format (n = 8 per group). Statistics for the effects of DSF intervention represent the p-value from a one-way ANOVA with Dunnett’s post hoc tests. *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001. e Clustering of liver proteins significantly impacted by DSF treatment as provided by String protein–protein interaction database. High confidence interaction score was selected (0.7). For reasons of clarity, some of the proteins present in the redox & drug metabolism cluster were labeled ‘a’ to ‘m’. These are: a, Aldh1a1; b, Ugdh; c, Cyp2b1; d, Cyp2e1; e, Cyp4a1; f, Cyp3a1; g, Ugt1a1; h, Ugt2b1; i, Cyp2c6v1; j, Cyp2c13; k, Cyp1a2; l, Cyp2c12; m, Gsta2. f Top enriched ten pathways generated from these experimental and predicted interactions map for “metabolic pathways,” “Redox and drug metabolism,” Tryptophan metabolism,” and “fatty acid (FA) and branched-chain amino acid (BCAA) catabolism.” g Gene-set enrichment analysis (GSEA) depicting a set of proteins involved in “metabolism of xenobiotics by CYP450” whose expression was significantly impacted by DSF supplementation. h PLS-DA of untargeted serum metabolomics (n = 8 in each group). i Heat map illustrates the relative average of each metabolite contributing to the group separation between control and DSF (low and high). j Relative abundance of a select group of metabolites. The data are represented in box and whisker plot format. *P < 0.05; **P < 0.01; ****P < 0.0001 vs. control group. Related to Supplementary Fig. 3 and Supplementary Tables 4 and 5.

Fig. 4. Disulfiram activates autophagy in cultured cells.

a Effect of disulfiram (DSF) on the degradation of long-lived proteins. Murine NIH3T3 fibroblasts in culture were labeled with 3H-leucine for 48 h and incubated in complete (serum+) or serum-free medium in the presence of 20 μM DSF or not. Rate of proteolysis was calculated as the percentage of the initial acid-precipitable radioactivity (proteins) transformed into acid-soluble radioactivity (amino acids and small peptides) at the indicated times. All values are mean ± SEM of three independent experiments, each performed with triplicate wells per time point. There is no error bar in the top panel because the error bar is shorter than the size of the symbol. *P < 0.05; **P < 0.01; ***P < 0.001 vs. without DSF. b–d Effect of DSF on basal and inducible autophagy. NIH3T3 cells expressing the tandem reporter mCherry-GFP-LC3 were exposed to 50 μM DSF for 24 h in complete (+Serum) or serum-free medium (−Serum). b Representative images of the individual and merged channels in cells where nuclei were highlighted by DAPI staining; c The number of autophagic vacuoles (AV), autophagosomes (APG), and autolysosomes (AUT) was determined by high-content microscopy. d Dose dependence of the activating effect of DSF on macroautophagy was analyzed in cells exposed to increasing concentrations of DSF. Number of AV (top panel), APG (middle panel), and AUT (bottom panel) was determined by high-content microscopy. e–g Structure activity of dithiocarbamate analogs on autophagy. e Structures of the eight compounds tested. f mCherry-GFP-LC3 reporter-expressing NIH3T3 cells were exposed for 24 h to 100 μM of DSF (A1) and DSF analogs (A2–A8) for the measure of AUT (top panel) or APG (bottom panel). g Changes in the number of AUT after exposure to increasing concentrations of A1, A3, A5, and A7. c–g Unless otherwise indicated, all values are mean ± SEM and quantifications were done in at least 2500 cells per condition in three different experiments using high-content microscopy. Abbreviations: A1, bis(diethylthiocarbamate) disulfide aka disulfiram; A2, ammonium pyrrolidine dithiocarbamate; A3, Mn2+-Zn2+ ethylenebis(dithiocarbamate); A4, Mn2+ ethylenebis(dithiocarbamate) aka pestanal; A5, Na+ diethyldithiocarbamate trihydrate; A6, triethylammonium N-(3,4-dichlorophenyl) dithiocarbamate; A7, S-cyanomethyl-N-methyl-N-(pyridin-4-yl) dithiocarbamate; A8, Zn2+ dimethyldithiocarbamate. Related to Supplementary Fig. 4.