The peroxisomal transporter ABCD3 plays a major role in hepatic dicarboxylic fatty acid metabolism and lipid homeostasis

Abstract

Peroxisomes metabolize a specific subset of fatty acids, which include dicarboxylic fatty acids (DCAs) generated by ω-oxidation. Data obtained in vitro suggest that the peroxisomal transporter ABCD3 (also known as PMP70) mediates the transport of DCAs into the peroxisome, but in vivo evidence to support this role is lacking. In this work, we studied an Abcd3 KO mouse model generated by CRISPR-Cas9 technology using targeted and untargeted metabolomics, histology, immunoblotting, and stable isotope tracing technology. We show that ABCD3 functions in hepatic DCA metabolism and uncover a novel role for this peroxisomal transporter in lipid homeostasis. The Abcd3 KO mouse presents with increased hepatic long-chain DCAs, increased urine medium-chain DCAs, lipodystrophy, enhanced hepatic cholesterol synthesis and decreased hepatic de novo lipogenesis. Moreover, our study suggests that DCAs are metabolized by mitochondrial fatty acid β-oxidation when ABCD3 is not functional, reflecting the importance of the metabolic compartmentalization and communication between peroxisomes and mitochondria. In summary, this study provides data on the role of the peroxisomal transporter ABCD3 in hepatic lipid homeostasis and DCA metabolism, and the consequences of peroxisomal dysfunction for the liver.

Keywords: dicarboxylic acids; lipid homeostasis; liver; mitochondria; peroxisome.

Figure 1.. Characterization of a second Abcd3 KO mouse model

A) Immunoblot using an antibody against ABCD3 in WT and Abcd3 KO liver extracts. Alpha-tubulin (α-tub) was used as the loading control. B) Bile acid analysis (in μM) in plasma of WT and Abcd3 KO fed mice (n=5). DHCA: Dihydroxycholestanoic acid; THCA: Trihydroxycholestanoic acid; CA: cholic acid; MCA: muricholic acid; DCA: deoxycholic acid; t-: taurine-. C) Representative images of WT and Abcd3 KO livers (WT fed n=4, Abcd3 KO fed n=5, WT fasted n=5, Abcd3 KO fasted n=7) with corresponding quantification of liver-to-BW (in %). BW: body weight. The effects in a two-way ANOVA were indicated as follows: G: Genotype. F: Feeding. I: Interaction. D) Representative images of intracellular hepatic glycogen evaluation with PAS staining in WT and Abcd3 KO fed and fasted livers. Scale bar = 100 μm. E) Immunoblot analysis using antibodies against ACOX1, EHHADH, CROT, CPT2, MCAD and CYP4A10 in liver homogenates from WT (n=5) and Abcd3 KO (n=5) fed mice. Alpha-tubulin (α-tub) was used as loading control. EHHADH and CYP4A10 were detected on the same membrane. ACOX1 and HMGCR from Fig. 4C were detected on the same membrane. CPT2 and MVK from Fig. 4C were detected on the same membrane. MCAD and MLYCD from Fig. 4E were detected on the same membrane. Individual values, the average and the standard deviation are graphed. * p<0.05; ** p<0.01; *** p<0.001 (unpaired, two-tailed Student’s t-test in B, and F; two-way ANOVA in C).

Figure 2.. Untargeted metabolomic analysis of WT and Abcd3 KO mouse livers

A) 2-D PCA scores plot between select PCs obtained from WT and Abcd3 KO hepatic metabolome (n=7). The explained variances are shown in brackets. B) Important metabolites selected by volcano plot with fold-change threshold (x-axis) = 2 and p-value threshold (y-axis) = 0.05. Blue and red circles represent metabolites above the p-value threshold, and below (blue) or above (red) the fold change threshold, respectively. Both fold-changes and p-values are log transformed. The top 5 metabolites below and above the fold-change threshold are indicated. C) Summary of pathway enrichment analysis using significantly altered metabolites with a KEGG ID. Scatterplot represents unadjusted p-values from integrated enrichment analysis and impact values from pathway topology analysis. The node color is based on the p-values and the node radius represents the pathway impact values. The 3 significantly altered KEGG pathways using Fisher’s exact t-test (adj. p < 0.1) are indicated.

Figure 3.. Changes in fatty acid metabolism in Abcd3 KO mice

A) Urinary C6-DCA, C8-DCA, C10-DCA, and C12-DCA (in mmol/mol creatinine) in WT (n=4 males, n=4 females) and Abcd3 KO (n=6 males, n=5 females) fed mice. B) Liver dicarboxylylcarnitine profile (in pmol/mg tissue) in WT (n=5 fed, n=7 fasted) and Abcd3 KO (n=5 fed, n=7 fasted) mice. Dicarboxylylcarnitine species that could not be distinguished from other hydroxyacylcarnitine species with the same nominal mass (isobaric compounds) are marked with an asterisk, as indicated here: C8DC*/C12-OH; C10DC*/C14-OH; C12DC*/C16-OH; C14DC*/C18-OH. C) Liver medium- and long-chain acylcarnitine profile (in pmol/mg tissue) in WT (n=5 fed, n=7 fasted) and Abcd3 KO (n=5 fed, n=7 fasted) mice. D) Measured [U-¹³C]-labeled C8-DC-carnitine, and C10-DC-carnitine (in pmol/mg of tissue) in mouse liver slices after 4-hr incubation of WT and Abcd3 KO mouse liver slices (n=4) with [U-¹³C]-C12-DCA alone or with [U-¹³C]-C12-DCA + L-aminocarnitine (L-AC). E) Plasma medium- and long-chain acylcarnitine profile (in μmol/L) in WT (n=5) and Abcd3 (n=6) KO mice treated with L-AC and subjected to food withdrawal during the photophase. Individual values, the average and the standard deviation are graphed.* p<0.05; ** p<0.01; *** p<0.001 (two-way ANOVA in A-D; unpaired, two-tailed student t-test in E). The effects in the two-way ANOVA are indicated as follows; G: Genotype; S: Sex; F: Feeding; L: L-AC; I: Interaction.

Figure 4.. ABCD3 deficiency alters hepatic cholesterol synthesis and de novo lipogenesis

A) Cholesterol synthesis (in “% of newly made”) was calculated from the incorporation of deuterium isotopes (²H) into cholesterol in the liver of WT (n=6 males and n=6 females) and Abcd3 KO (n=6 males and n=6 females) mice. B) Total cholesterol amount (in “μmol/g tissue”) in the liver of WT (n=6 males and n=6 females) and Abcd3 KO (n=6 males and n=6 females) mice was measured by mass spectrometry. C) Immunoblot using antibodies against HMGCR and MVK in liver homogenates from WT (n=5) and Abcd3 KO (n=5) fed mice. Alpha-tubulin (α-tub) was used as loading control. D) De novo lipogenesis (DNL, in “% of newly made”) was calculated from the incorporation of deuterium isotopes (²H) into the corresponding TG-bound fatty acids (palmitate, oleate, and stearate) in the liver of WT (n=6 males and n=6 females) and Abcd3 KO (n=6 males and n=6 females) mice. E) TG-bound fatty acid (palmitate, oleate, and stearate) content (in “μmol/g tissue”) was measured by mass spectrometry. F) Immunoblot using antibodies against phosphorylated ACC on the residue Ser79 [pACC (S79)] and MLYCD in liver homogenates from WT (n=5) and Abcd3 KO (n=5) fed mice. Total ACC or alpha-tubulin (α-tub) were used as loading controls as indicated in the x-axis of the graph. Individual values, the average and the standard deviation are graphed.* p<0.05; ** p<0.01; *** p<0.001 (unpaired, two-tailed student t-test).