Major Depressive Disorder is Associated with Impaired Mitochondrial Function in Skin Fibroblasts

Abstract

Mitochondrial malfunction is supposed to be involved in the etiology and pathology of major depressive disorder (MDD). Here, we aimed to identify and characterize the molecular pathomechanisms related to mitochondrial dysfunction in adult human skin fibroblasts, which were derived from MDD patients or non-depressive control subjects. We found that MDD fibroblasts showed significantly impaired mitochondrial functioning: basal and maximal respiration, spare respiratory capacity, non-mitochondrial respiration and adenosine triphosphate (ATP)-related oxygen consumption was lower. Moreover, MDD fibroblasts harbor lower ATP levels and showed hyperpolarized mitochondrial membrane potential. To investigate cellular resilience, we challenged both groups of fibroblasts with hormonal (dexamethasone) or metabolic (galactose) stress for one week, and found that both stressors increased oxygen consumption but lowered ATP content in MDD as well as in non-depressive control fibroblasts. Interestingly, the bioenergetic differences between fibroblasts from MDD or non-depressed subjects, which were observed under non-treated conditions, could not be detected after stress. Our findings support the hypothesis that altered mitochondrial function causes a bioenergetic imbalance, which is associated with the molecular pathophysiology of MDD. The observed alterations in the oxidative phosphorylation system (OXPHOS) and other mitochondria-related properties represent a basis for further investigations of pathophysiological mechanisms and might open new ways to gain insight into antidepressant signaling pathways.

Keywords: adenosine triphosphate; bioenergetics; calcium imaging; major depression; mitochondria; mitochondrial DNA copy number; mitochondrial membrane potential; oxidative phosphorylation; skin fibroblasts.

Figure 1

Oxygen consumption rates (OCR) (A–G) and extracellular acidification rate (ECAR) (G) measured by Seahorse XFp Flux Analyzer in major depressive disorder (MDD, indicated in orange) and control (indicated in grey) fibroblasts under non-treated conditions, as well as after one week of dexamethasone (DEX, 1 µM) or galactose (GAL) stress (10 mM GAL, glucose-free). The Mito Stress Test allows the analysis of basal (A) and maximal respiration (B), as well as oxygen consumption related to ATP production (C). Bar graphs show normalized mean OCR values + standard error of the mean (SEM); MDD n = 16, control n = 16. Significant differences between MDD and non-depressive controls are indicated with *. Mito Stress Test in fibroblasts (D–F). Exemplary curves for OCR measurement during the Mito Stress Test for MDD and control fibroblast lines (pair #6). Sequential injection of oligomycin, carbonyl cyanide-4-(trifluoromethoxy)phenylhydrazone (FCCP) and rotenone/antimycin A affects OCR by interaction with the electron transport chain (ETC) complexes ETC in the oxidative phosphorylation system (OXPHOS) of untreated (D), DEX-treated (1 µM, 7 days) (E) and GAL-stressed fibroblasts (glucose-free, 10 mM galactose, 7 days) (F). (G) Energy map of fibroblasts. Mean OCR in dependence of mean ECAR are shown for MDD and control fibroblasts for non-treated, DEX-treated (1 µM, 7 days) and GAL-stressed (glucose-free, 10 mM galactose, 7 days) conditions. Significant effects of treatment were found for DEX (# p ≪ 0.05, compared with non-treated, analysis of variance (ANOVA) with repeated measures, Greenhouse–Geisser correction, post-hoc analysis with Bonferroni) and GAL (§ p ≪ 0.05, compared with non-treated, ANOVA with repeated measures, Greenhouse–Geisser correction, post-hoc analysis with Bonferroni), Data are shown as mean OCR ± SEM vs. mean ECAR ± SEM; MDD n = 16, control n = 16.

Figure 2

ATP content in fibroblasts. ATP content in MDD and control fibroblasts under non-treated, DEX-treated (1 µM, 7 days) or GAL-stressed conditions. Significant differences were found for MDD vs. control, non-treated (* p < 0.05, Wilcoxon matched-pairs signed rank test) and DEX (** p < 0.01, compared with control, Wilcoxon matched-pairs signed rank test). Bar graphs show normalized mean RLU values ± SEM. Dots show the distribution of single RLU values for MDD and control fibroblast lines; MDD n = 16, control n = 16.

Figure 3

Mitochondrial membrane potential of fibroblasts. (A) Red/green (JC-1 aggregate/monomer) ratios of MDD and control fibroblasts under non-treated, DEX-treated (1 µM, 7 days), or GAL-stressed conditions (glucose-free, 10 mM galactose, 7 days). Significant differences were found for MDD vs. control, non-treated (* p < 0.05, compared with control, Student’s t-test, paired, two-tailed). Bar graphs show mean red/green ratios ± SEM, MDD n = 16, control n = 16). Dots show the distribution of single red/green values for MDD and control fibroblast lines under the indicated conditions; MDD n = 16, control n = 16. (B) Fluorescence microscopy image of fibroblasts loaded with the cationic dye JC-1 under basal conditions (left) and after treatment with 20 µM FCCP to uncouple the proton gradient leading to depolarization of the MMP. Aggregates of the dye fluoresce red, monomers fluoresce green. Scale bar indicates 20 µM.

Figure 4

Cytosolic Ca²⁺ homeostasis in fibroblasts. Shown are the Fura-2 340 nm/380 nm ratios of MDD and non-depressed control fibroblast lines under non-treated, DEX-treated (1 µM, 7 days) or GAL-stressed conditions (glucose-free, 10 mM galactose, 7 days). No significant differences were found. Bar graphs show mean ratios (340 nm/380 nm; ratios ± SEM). Dots show the distribution of ratios (340 nm/380 nm) for MDD and control fibroblast lines under indicated conditions (MDD n = 16, control n = 16).

Figure 5

Mitochondrial DNA (mtDNA) copy number per nDNA in 16 MDD patient and control fibroblast cell lines. No significant differences were found. Bar graph show mean mtDNA copy number ± SEM. Dots show the single values of mtDNA copy numbers for MDD and control fibroblast lines; MDD n = 16, control n = 16.