Mitochondrial gene signature in the prefrontal cortex for differential susceptibility to chronic stress

Weger, Meltem; Alpern, Daniel; Cherix, Antoine; Ghosal, Sriparna; Grosse, Jocelyn; Russeil, Julie; Gruetter, Rolf; de Kloet, E. Ronald; Deplancke, Bart; Sandi, Carmen

doi:10.1038/s41598-020-75326-9

Download PDF

Article
Open access
Published: 27 October 2020

Mitochondrial gene signature in the prefrontal cortex for differential susceptibility to chronic stress

Meltem Weger^1,5,
Daniel Alpern^2,6,
Antoine Cherix³^nAff7,
Sriparna Ghosal¹,
Jocelyn Grosse¹,
Julie Russeil²,
Rolf Gruetter³,
E. Ronald de Kloet⁴,
Bart Deplancke^2,6 &
…
Carmen Sandi¹

Scientific Reports volume 10, Article number: 18308 (2020) Cite this article

5128 Accesses
36 Citations
28 Altmetric
Metrics details

Subjects

Abstract

Mitochondrial dysfunction was highlighted as a crucial vulnerability factor for the development of depression. However, systemic studies assessing stress-induced changes in mitochondria-associated genes in brain regions relevant to depression symptomatology remain scarce. Here, we performed a genome-wide transcriptomic study to examine mitochondrial gene expression in the prefrontal cortex (PFC) and nucleus accumbens (NAc) of mice exposed to multimodal chronic restraint stress. We identified mitochondria-associated gene pathways as most prominently affected in the PFC and with lesser significance in the NAc. A more detailed mitochondrial gene expression analysis revealed that in particular mitochondrial DNA-encoded subunits of the oxidative phosphorylation complexes were altered in the PFC. The comparison of our data with a reanalyzed transcriptome data set of chronic variable stress mice and major depression disorder subjects showed that the changes in mitochondrial DNA-encoded genes are a feature generalizing to other chronic stress-protocols as well and might have translational relevance. Finally, we provide evidence for changes in mitochondrial outputs in the PFC following chronic stress that are indicative of mitochondrial dysfunction. Collectively, our work reinforces the idea that changes in mitochondrial gene expression are key players in the prefrontal adaptations observed in individuals with high behavioral susceptibility and resilience to chronic stress.

The serotonin theory of depression: a systematic umbrella review of the evidence

Article Open access 20 July 2022

Joanna Moncrieff, Ruth E. Cooper, … Mark A. Horowitz

Gene expression networks regulated by human personality

Article Open access 04 March 2024

Coral del Val, Elisa Díaz de la Guardia-Bolívar, … Claude Robert Cloninger

A concerted neuron–astrocyte program declines in ageing and schizophrenia

Article Open access 06 March 2024

Emi Ling, James Nemesh, … Steven A. McCarroll

Introduction

Chronic stress is a major risk factor for the development of psychopathologies, such as depression^1,2,3, a severe mental disease with a high prevalence and limited effective treatment possibilities^4,5. Despite extensive clinical and preclinical research efforts carried out to date (e.g.,^{6,7,8,9,10,11,12}), there is still a lack of understanding about the specific biological changes that link chronic stress with depressive symptoms.

Brain regions, such as the prefrontal cortex (PFC) and the nucleus accumbens (NAc) that are implicated in depression pathogenesis^{11,13,14,15,16} show high susceptibility to display structural and functional alterations under exposure to chronic stress^12,17,18,19. In the search for potential mechanisms underlying those changes, several studies have reported gene expression changes in the PFC and NAc in both, depressive patients^16,20,21,22 and rodent models of chronic stress^23,24,25. A very recent comparative study has reported considerable overlap between transcriptional signatures in the PFC and NAc in human major depressive disorder and in three mouse chronic stress models, validating the use of these animal approaches to model depression²⁶. Interestingly, only the enrichment of genes involved in mitochondrial function in both PFC and NAc was commonly shared between human depression and every mouse model, whereas for all other shared gene pathways each mouse model captured distinct aspects of major depression abnormalities²⁶. While these findings suggest that alterations in brain mitochondrial-related gene expression may be a critical evolutionarily conserved pathway linking chronic stress with depression, this study notably did not follow up on the mitochondrial data²⁶.

Mitochondria are multifunctional life-sustaining organelles responsive to life stress and increasingly recognized as potential intersection points between psychosocial experiences and biological stress responses^27,28. For example, stress-induced mitochondrial impairments leading to decreased ATP synthesis and/or excessive production of reactive oxygen species (ROS) may affect behavior by interfering with neuroplasticity and causing neurotoxicity^28,29,30. Indeed, mitochondrial dysfunction is emerging as a vulnerability factor for human depression^31,32 and stress-related psychopathologies^28,30. However, information regarding the precise stress-induced changes in mitochondria-associated genes in brain regions relevant for depression symptomatology, and how they relate to depression-related behavioral manifestations, is scarce. Only a few studies have investigated alterations in the transcriptional expression of mitochondria-associated genes in the brain of major depression disorder (MDD) subjects³³, including the PFC³⁴. In rodent models of chronic stress, a recent multi-omics study including genome-wide analyses identified mitochondrial-related alterations in gene expression in the bed nucleus of stria terminalis, an anxiety-related brain region³⁵.

Here, we performed a genome-wide transcriptomic study using bulk transcriptomics to assess and define mitochondrial gene expression signatures in the PFC and NAc upon chronic stress. We firstly investigated whether mitochondrial gene sets and pathways would show enriched differential expression in the PFC and NAc between mice submitted to multimodal chronic restraint stress (mCRS) and unstressed control (CTR) mice. As our results revealed prominent chronic stress-induced alterations in mitochondrial pathways, we then performed a detailed analysis to understand to which extent nuclear (nDNA) and mitochondrial DNA (mtDNA)-encoded mitochondria-related genes were affected in these brain regions. In addition, to assess whether the detected mitochondrial gene expression changes are a chronic stress feature that generalizes to other protocols, we reanalyzed a previously published gene expression data set of PFC and NAc from chronic variable stress (CVS) mice²⁰. Moreover, to provide translational implications of our findings, we compared our data with differential gene expression in major depression disorder (MDD) subjects²⁰. Lastly, we provide evidence that the here defined mitochondrial gene expression signature reflects the mice behavioral profiles and their susceptibility to chronic stress.

Results

The mCRS protocol leads to changes in physiology and behavior

Mice were randomized into CTR or mCRS groups following matching according to their body weight and anxiety-like behaviors (i.e., percent time in the open arms of the EPM; see experimental schedule in Fig. 1A). The OF and NO tests confirmed that there were no a priori group differences in anxiety, locomotion, and exploration (Fig. S1A,B). During exposure to mCRS, stressed mice showed reduced body weight gain and lower food intake at the beginning of the stress paradigm than CTRs (Fig. 1B). Stress also led to increased basal blood CORT levels, as measured four days after the end of the mCRS protocol (Fig. 1C) and increased adrenal glands’ weight (Fig. 1D).

At the behavioral level, and as compared to CTRs, mice exposed to mCRS showed: (1) decreased sociability, as indicated by their reduced time spent in the social interaction zone and increased social avoidance score (Fig. 1E); (2) increased immobility in the FST (Fig. 1F); and (3) unaltered saccharin preference (Fig. 1G). Importantly, the computed integrated z-score indicated that mCRS exposed animals showed a significant alteration when the performance across the different tested behavioral domains was considered (Fig. 1H).

mCRS severely affects the PFC transcriptome and leads to changes in mitochondrial pathways

Next, we performed a hypothesis-free transcriptomic analysis to explore the effects of mCRS on PFC and NAc gene expression (see Table S1 for a list of differentially expressed genes in the PFC and NAc, and Fig. S2A, B for mapping efficiency). The overall number of differentially expressed genes in the PFC following mCRS exposure was higher than the number of differentially expressed genes in the NAc (Fig. 2A–C). The top 20 differentially expressed genes contained genes previously implicated in stress-vulnerability and/or depression (for details, see Fig. S2C,D), indicating that mCRS leads to pathological changes in the PFC and NAc transcriptomes. We then performed a gene set enrichment analysis to identify the most affected pathways in the PFC and NAc (Table S2) upon mCRS. The top 10 enriched pathways in the PFC, interestingly, included gene pathways implicated in mitochondrial energy synthesis, such as “oxidative phosphorylation”, “aerobic electron transport chain”, and “mitochondrial electron transport” (Fig. 2D). In line with these findings, translation (e.g., “ribosome” and “cytoplasmic translation”) as a cellular energy-depending process was also affected in the PFC (Fig. 2D). Gene pathways implicated in mitochondrial energy synthesis were also detected in the NAc (e.g., “oxidative phosphorylation”, “energy coupled proton transport”; Fig. 2E), although with a lower significance than in the PFC. These observations highlight profound transcriptional imbalances in mitochondrial gene expression upon mCRS particularly in the PFC.

Gene expression profiles of mitochondria-associated genes following mCRS

Our former analyses pointed to changes in mitochondrial pathways upon mCRS in the PFC, and to some extent in the NAc. We next focused on the mCRS-induced changes in mitochondria-associated genes to decipher the contribution of nDNA and mtDNA encoded genes. mCRS led to an up-regulation of mitochondria-associated genes encoded by both nDNA and mtDNA encoded genes in the PFC (Fig. 3A). A deregulation of mitochondria associated genes was not detected in the NAc (Fig. 3A). Specifically, all of the detected mtDNA encoded genes of the oxidative phosphorylation (OXPHOS) complexes I, III and IV (i.e., mt-Nd1, mt-Nd2, mt-Nd4, mt-Nd5, mt-Nd6, mt-Cytb, and mt-Co1) were up-regulated in the PFC of mCRS mice (Fig. 3B,C). Note that the other six mtDNA encoded genes of the OXPHOS complexes (i.e., mt-ND3, mt-ND4L, mt-Co2, mt-Co3, mt-Atp6, and mt-Atp8) were too low expressed to be reliably detected in our study. We verified the expression changes of selected mtDNA encoded genes by qRT-PCR and confirmed upregulation of the complex I genes mt-Nd1 and mt-Nd-3 (Fig. S3). To obtain information about the expression of complex V genes, we analyzed mt-Atp6 expression and detected its up-regulation in the PFC of stressed mice as well (Fig. S3).

To evaluate whether the here detected mitochondrial gene expression changes are a chronic stress feature that generalizes to other chronic stress protocols, we also tested and showed that the up-regulation of both nDNA and mtDNA encoded genes is not solely specific to mCRS but can be also detected following CVS exposure (Fig. 3A, Fig. S4).

Furthermore, in a translational effort, we next tested whether the detected mitochondrial gene expression changes in chronically stressed mice are also present under conditions of MDD. For this end, we reanalyzed the data of post-mortem brain tissue from MDD patients and matched control subjects previously published by Labonte et al.²⁰. For our comparison, we used the data from vmPFC (henceforth termed PFC) and NAc. Interestingly, there was a high overlap in mitochondria-associated gene expression between the data in mice and MDD subjects for mtDNA-encoded genes in the PFC (Fig. 3A,B). Therefore, altogether these analyses point to mtDNA encoded gene expression changes as particularly susceptible to be altered in the context of chronic stress and depression.

Chronic stress leads to changes in mitochondrial function and metabolism

The transcriptional changes in mitochondria associated genes indicated the presence of mitochondrial dysfunctions in mCRS mice. Thus, we assessed whether mCRS impacts mitochondrial function in the PFC and the NAc by addressing mitochondrial respiration capacity ex vivo. Our results showed a reduction in mitochondrial respiration capacity in mCRS mice in the PFC (Fig. 4A) but no changes in the NAc (Fig. 4B). To gain more insights on the impact of mCRS on brain metabolism, we performed ¹H-magnetic resonance spectroscopy (¹H-MRS; Fig. S5) and evaluated energy-related (i.e., creatine, phospho-creatine, glucose) metabolites, including glycolytic (i.e., alanine, lactate) ones³⁶. Levels of glucose (Glc), the brain’s primary energy source, were decreased in the PFC of mCRS mice (Fig. 4C). No mCRS-induced changes in metabolites were detected in the NAc (Fig. 4D).

Next, we tested whether mCRS induces the mitochondrial unfolded stress-response (UPR^mt) which is initiated by the transcription factors ATF-5, ATF-4 and CHOP/DDIT3^37,38,39 upon mitochondrial dysfunctions to maintain homeostasis⁴⁰. To predict in silico for potential transcription factor binding sites and thus activity of ATF-5, ATF-4 and CHOP/DDIT3 (“predicted motif activity”) in the promoters of differentially expressed genes in the PFC and NAc upon mCRS, we scanned ± 500 bp around the annotated transcription start site of a gene. Interestingly, our analysis revealed that the predicted ATF-4 and CHOP/DDIT3 motif activity is higher in the PFC of mCRS mice compared to CTR animals (Fig. 4E). In contrast, no increased predicted motif activity of any of the tested transcription factors was detected in the NAc of mCRS mice (Fig. 4F), indicating that the UPR^mt via the ATF4 pathway due to mitochondrial dysfunctions is active in the PFC of mice following mCRS.

mtDNA encoded gene expression in the PFC correlates with mice behavioral profiles

To gather more information about whether the expression of mitochondria-associated genes in the PFC of mCRS mice can be related to mice behavioral profiles, we applied a model selection analysis of gene expression that considers the individual differences in stress and behavioral responses of the tested mice. Specifically, this method allows to extract genes either better correlating with 1) the treatment itself (i.e., mCRS) or with 2) the consequences of the treatment on behavior (i.e., z-score). In our analysis, the glucocorticoid-dependent mTOR regulator Ddit4⁴¹ would be an example for genes of group 1 and the mitochondrial gene Ndufv2 would represent genes of group 2 (Fig. 5A, Table S1).

A gene set enrichment analysis of gene pathways better correlating with the behavioral profiles of mCRS mice led to the observation that “oxidative phosphorylation” is one of the pathways that is better explained by individual behavioral changes (z-score) than simply the absence or presence of mCRS (Fig. 5B, Table S2). A more detailed analysis of the OXPHOS complex genes showed that genes of complex I, but not of the other complexes, correlate better with the integrated behavioral z-score than with mCRS (Fig. 5C). In fact, in our study all detected mtDNA encoded genes correlated better with the z-score than with mCRS (Fig. 5D,E). In addition, we reevaluated our data by separating mCRS animals according to their susceptibility to stress [i.e., high- (HS) versus low-susceptible (LS), in accordance with⁴²]. This approach showed that mCRS causes increased expression of mtDNA encoded genes only in HS but not in LS animals (Fig. 5F).

Discussion

Mitochondrial dysfunction has been postulated as a key etiopathology factor for the development of chronic stress-induced psychopathological alterations, including depression^28,30. Neural adaptations to chronic stress are energetically costly and mitochondria are ideally positioned to contribute to them by providing energy and stress signaling molecules^28,30. However, systematic studies describing the precise alterations in mitochondria-associated genes in the brain following exposure to chronic stress remain scarce. In this study, using a genome-wide hypothesis-free transcriptomic approach, we identified mitochondria-associated gene pathways among the categories showing the most prominent responses in the PFC and to a lesser extent in the NAc following chronic stress exposure in mice. Most importantly, we revealed that mtDNA genes coding for different subunits of the OXPHOS complexes in the PFC as particularly affected by stress, and showed that their upregulation is particularly observed in stress-vulnerable mice.

First, we confirmed that our multimodal chronic restraint stress protocol leads to reduced sociability and increased passive coping responses under adversity, two behavioral alterations also typically observed in depressive individuals^43,44,45. Then, the computation of the number of genome-wide stress-responsive genes, including mitochondria-associated ones, pointed at a higher reactivity of the PFC to stress than the NAc. This observation is in line with other reports showing that, at the transcriptional level, the PFC is more severely affected than the NAc by chronic stress or major depression^20,24,46. Rodent studies have reported opposite effects of chronic stress in neuronal structure in these two brain regions, frequently indicating decreases in spine density and dendritic complexity in the PFC^47,48,49,50 while increase in the NAc^49,51 (but see⁵²). However, a note of caution should be added here, not only because the evidence is mixed as indicated above, but also because different cell types are involved in the referred studies. While data from the NAc has been obtained from the analysis of GABAergic projection neurons (i.e., medium spiny neurons), the atrophy described in the PFC refers to excitatory pyramidal cells. In fact, opposite results (i.e., dendritic hypertrophy) have been found in the medial PFC when analyzing the effects of chronic stress in a subpopulation of interneurons⁵³. However, the relatively lower susceptibility of the NAc to show transcriptomic changes following chronic stress, including mitochondria-associated ones, contrasts with the highest vulnerability shown by the NAc than the PFC to display changes in neurometabolic markers following stress¹⁸. A possible reason for this discrepancy may be the different stress protocols used (i.e., 10-days chronic social defeat¹⁸ vs. the here applied 3-week mCRS). Note, however, that a 3-way chronic restraint protocol results as well in major metabolic changes in the NAc in stress-vulnerable mice⁵⁴. In the future, it will be important to integrate transcriptomic and metabolic analyses in these two brain regions, ideally at the cell-type level, to further our understanding of stress-induced metabolic regulation at the transcriptional level.

Mitochondria are multifunctional life-sustaining organelles that have as a major function to perform OXPHOS, the metabolic pathway in which the electron transport chain, through its complexes I–IV, exchanges protons and electrons to eventually produce adenosine triphosphate (ATP) by complex V. Subunits of the mitochondrial complexes of OXPHOS are encoded not only by nDNA but also by mtDNA²⁸. Our mitochondria-focused analysis showed that mCRS mice exhibited an up-regulated expression of mtDNA encoded genes of virtually all OXPHOS complexes, except for complex II. We furthermore showed that mtDNA encoded genes are also up-regulated in the PFC of mice subjected to another chronic stress paradigm²⁰. These results align well with proteomic data indicating chronic stress-induced changes in the levels of proteins involved in mitochondrial transport and OXPHOS in the PFC⁵⁵. Likewise, chronic stress was also shown to lead to up-regulated expression of the mtDNA-encoded complex I gene mt-ND6 in the rat hippocampus⁵⁶. Altogether, these findings, strongly supported the notion that an upregulated expression of mtDNA encoded genes is a common response to chronic stress. Importantly, mouse gene deletion approaches involving genes of the mitochondrial OXPHOS complexes were shown to lead to alterations in physiological stress responses^27,57.

Importantly, we found stress-induced increases in the expression of genes coding for subunits of different complexes of the OXPHOS machinery in the PFC along with decreased levels of the energy metabolite glucose and reduced maximal capacity of mitochondrial respiration as evaluated ex vivo. Diminished mitochondrial respiration in the PFC was also reported for other chronic stress models⁵⁸ and in fibroblasts from major depressive disorder patients⁵⁹. Glucose levels are also typically found decreased in the brain of depressed patients^60,61,62 and following chronic stress in rodents⁶³, in association with a reduction in glucose transporters expression⁶³. Given that glucose is the main source of energy for the mammalian brain⁶⁴, through a key contribution to OXPHOS processes⁶⁵, the observed increases in genes coding for OXPHOS complexes subunits may be an allostatic response to compensate for reduced fuel^28,66,67. However, given that our neurobiological endpoints were only measured after chronic stress exposure, we cannot discard that they are not a reaction to stress-induced enhancements in mitochondrial function potentially observed at earlier time points^56,68, particularly in the PFC⁶⁹. Eventually, the reported compensatory changes could represent responses to the “allostatic load” induced by chronic stress⁷⁰. Such a view would reconcile the reductions in dendrite complexity and spine density typically observed following chronic stress in the PFC^47,48 and the increased neuronal vulnerability triggered by elevated rate of mitochondrial OXPHOS^71,72. Increased rate of mitochondrial OXPHOS would lead to enhanced production of ROS making neurons more vulnerable and less capable of sustaining complex axonal arborization^71,73. This view is further reinforced by the fact that the increased expression of mtDNA encoded OXPHOS genes upon chronic stress relates with the behavioral profiles of the mice and was particularly observed in high-susceptible animals. Altogether, our data point to that mCRS leads to profound changes in mitochondrial outputs in the PFC that are indicative for mitochondrial dysfunction in this brain region.

Our findings of high transcriptional susceptibility of mitochondrial pathways to stress align well with the recent observation of similar marked dysregulations across three mouse models of chronic stress and in human MDD both in the PFC and in the NAc²⁶. Here, we go beyond those observations and provide translational evidence for similar upregulation of mtDNA encoded OXPHOS genes in the PFC of subjects with major depression disorder, a new finding we made by reanalyzing a previously published transcriptome data set²⁰. Altogether, these findings reinforce the idea that changes in mitochondrial gene expression are key players in the prefrontal adaptations observed in individuals with high behavioral susceptibility to chronic stress.

In the future, it will be important to perform similar mitochondrial analyses as the ones included in the current study to investigate chronic stress effects in other brain regions critically involved in the brain response to chronic stress and MDD, such as the orbitofrontal cortex (OFC), hippocampus and amygdala^74,75. These studies will be particularly relevant to understand the link between mitochondria and stress-induced changes in neuronal structure as, in contrast to the hypotrophy induced by stress in the medial PFC^47,48,49,50 or in the hippocampus^76,77, the basolateral amygdala⁷⁸ (but see⁷⁹) and the OFC show dendritic hypotrophy⁸⁰ following similar stress protocols. In addition, future functional studies will be needed to elucidate the underlying molecular mechanisms of how the herein defined mitochondrial gene expression changes can act as a vulnerability factor for the development and/or pathology of stress-induced depression. Some evidence from the literature points to glucocorticoids via the glucocorticoid receptor (GR) as potential mediators. Previous studies have shown that both mitochondrial function⁸¹ and glucocorticoid actions^82,83 in the NAc is important for the establishment of social dominance in rats, a considered vulnerability factor for depression⁸⁴. Moreover, the GR implicated in the mesocorticolimbic dopamine system with chronic stress vulnerability^85,86, and showed that the GR has the capability to translocate into brain mitochondria^68,87,88 and to bind on the mitochondrial genome^56,89 in a glucocorticoid-dependent manner⁵⁶. It will also be important to test whether mitochondrial-targeted interventions known to ameliorate stress and anxiety responses^54,90,91,92 would reverse both the behavioral and mtDNA transcriptional changes observed in stress-vulnerable mice.

In summary, we here defined the mitochondrial gene signatures upon chronic stress in the PFC and NAc, two key brain regions implicated with stress-vulnerability and depression. Our data point to a higher susceptibility of the PFC to chronic stress on a transcriptional level than of the NAc. Moreover, our studies indicate that chronic stress induces a conserved response in mtDNA encoded gene expression in the PFC. Future studies will elucidate how these mitochondrial gene signatures might mediate differential susceptibility to chronic stress.

Material and methods

Animals

Experimental male 6-week old C57Bl6/J mice and retired old-breeder CD1 mice serving as social targets for the SIT test were purchased from Charles River Laboratories. Experimental animals were housed four per cage (except for the CD1 mice that were due to the experimental design under single-housed conditions) and were allowed to acclimate to the vivarium for at least one week. Animals were handled for at least 1 min per day for three subsequent days and then were weighted upon arrival and then weekly to ensure good health. Animals were kept under standard housing conditions on corn litter in a temperature- (23 ± 1 °C) and humidity- (40%) controlled animal room with a 12 h light/dark cycle (7 am–7 pm) with ad libitum access to food and water. All animal procedures were conducted in accordance with the regulations and approved by the Cantonal Ethics Committee of the Canton Vaud, Switzerland.

Behavioral and physiological characterization

Mice (7 weeks old) were randomized in two equivalent groups according to their body weight and performance in the elevated plus maze (EPM). They were then exposed at the age of 10 weeks to open field and novel object (OF + NO) tests to verify the effectiveness of the randomization. Starting at an age of 11 weeks one group of mice was then exposed to chronic multimodal chronic restraint stress (mCRS; n = 16), the other was kept as unstressed controls (CTR; n = 16). Body weight and food intake was measured daily before stress exposure. Mice (14 weeks old) were then tested in the social interaction test (SIT), the forced swim test (FST), and the saccharin preference test (SaPT). Trunk blood was collected during sacrifice (15 weeks old) for basal plasma corticosterone (CORT) measurements and the adrenal glands were dissected to determine adrenal glands’ weight.

EPM

The EPM was performed under baseline conditions to randomize animals into experimental groups according to their anxiety levels, a vulnerability factor for depression⁹³. The apparatus consisted of black PVC with a white floor with two open arms and two closed arms (30 × 5 cm) placed 65 cm above the floor. Light conditions during the testing were 12 lx in the open arms and 3–4 lx in the closed arms. Experimental animals were placed into the center of the EPM and exploratory behavior were analyzed for 5 min with a video-tracking software (Ethovision 11.0 XT, Noldus, Information Technology). The time an experimental animal spent in the open arms, center, and closed arms was measured.

OF + NO tests

The OF + NO tests were performed under light conditions of 7 lx. For the OF test, mice were placed in a white open field arena (50 × 50 cm) and were allowed to explore the arena for 10 min. For the NO test, experimental animals were then provided an empty water bottle as a novel object into the center of the arena, and were monitored for an additional time of 5 min. Via a video tracking system (Ethovision 11.0 XT, Noldus, Information Technology) the times an animal spend at the different zones of the arena (wall, intermediate, center) and the total distance travelled was calculated.

mCRS paradigm

The mCRS paradigm includes daily restraint of mice in 50 ml falcon tubes equipped with air holes under changing conditions. The order was as follows: (1) 2 h restraint (9 to 11 am) at 89 lx light; (2) 3 h restraint (10 am to 1 pm) in darkness; (3) 1 h restraint (3 to 4 pm) at 89 lx light and loud music (65–70 dB); (4) 2 h restraint (9 to 11 am) at 144 lx light; (5) 1 h restraint (4 to 5 pm) at 89 lx light and slight jostling on a shaker; (6) 2 h restraint (10 am to 12 pm) under stroboscope light; (7) 40 min restraint (5 to 5:40 pm) at 89 lx light in the presence of 2,3,5-Trimethyl-3-thiazoline (TMT) odor; 8) 2 h restraint (9 to 11 am) in darkness; 9) 2 h restraint (9 to 11 am) at 144 lx light and loud music (65–70 dB); (10) 3 h restraint (10 am to 1 pm) at 89 lx light; (11) 3 h restraint (1 to 3 pm) at 144 lx light; (12) 2 h restraint (9 to 11 am) at stroboscope light; (13) 30 min restraint (10 to 10:30 am) at 89 lx light in the presence of TMT; (14) 1 h restraint (4 to 5 pm) at 89 lx light and slight jostling on a shaker; (15) 3 h (10 am to 1 pm) at 89 lx light; (16) 2 h (2 to 4 pm) at 144 lx light; (17) 2 h restraint (9 to 11 am) at 89 lx light and loud music (65–70 dB); (18) 1 h restraint (3 to 4 pm) under stroboscope light; (19) 40 min restraint (1 to 1:40 pm) at 144 lx light in the TMT odor; (20) 1 h restraint (10 to 11 am) at 89 lx light and slight jostling on a shaker; (21) 2 h restraint (9 to 11 am) at 89 lx light. After 21 days, 30 min of restraint stress at 89 lx light was continued in the afternoon between the behavioral tests to avoid recovery of the animals from the stressor.

SIT

The SIT⁹⁴ consisting of two consecutive sessions of 2.5 min was performed in a white open field arena (42 × 42 cm) at red light conditions. During the first session, animals were allowed to explore the arena including an empty wire mesh. During the second session, a CD1 mouse separated with a wire mesh was introduced into the arena as a social target. With a video tracking system (Ethovision 11.0 XT, Noldus, Information Technology) the time an experimental animal spent in the interaction zone (24 × 14 cm) and in the corner zones (each 9 × 9 cm) was measured. Four wild-type animals had to be removed from the analysis due to technical reasons such as an early stopping of the video recording. The social avoidance score was calculated with the following parameters as described in⁹⁵: (1) the time spend in the interaction zone in the presence of the social target, (2) the time spend in the corner zones in the presence of the social target, (3) the social interaction ratio (time spent in the interaction zone in the presence of the social target/time spent in the interaction zone in the absence of the social target), (4) the corner zone ratio (time spent in the corner zones in the presence of the social target/time spent in the corner zones in the absence of the social target).

FST

The FST⁹⁶ was performed for 6 min under light conditions of 90 lx. Mice were placed in 5 L beakers with 3.5 L of tap water at a temperature of 23–25 °C. Immobility time (the absence of any movement except to keep the animal head above the water) during the last 4 min of the session was scored manually.

SaPT

For the SaPT, animals were transferred into new home cages with two drinking bottles containing drinking water to habituate them to the new set-up. The next day, one of the bottles was replaced with a 0.05% saccharin (Sigma, #S1002) solution and the weight of the water bottles was measured every day for three subsequent days. The bottles were interchanged after each weight measurement to avoid the development of side preference of the animals. Total saccharin consumption was calculated as a percentage of amount of saccharin solution consumed divided by the total volume consumed.

Behavioral z-score index

In order to have an integrated measure of behavioral phenotypes, we performed an integrated behavioral z-score composed of the integrated computation of respective z-normalization across behavioral tests. This approach allows standardizing observations obtained in different experiments, enhancing the reliability of the behavioral phenotyping and increasing analytical possibilities⁹⁷. Specifically, our z-score was calculated by averaging the z-normalized raw data [(x-mean of CTR group)/STD of CTR group] obtained per mouse in the SIT (social avoidance), FST (time of immobility), and SaPT (% preference) tests. The z-score was determined for mice for which the behavioral parameters could be collected for all tests (i.e., 11 CTR, 16 mCRS mice). In agreement with previous studies⁴², we subdivided chronically stressed mice into low-susceptible (LS) and high-susceptible (HS) clusters according to their z-score. Thus, mCRS mice which fell into the standard deviation from the mean of the control group were designated as LS, as they exhibit comparable behavior to unstressed CTRs, and animals that fell outside were designated as HS.

CORT levels

Trunk blood was collected (sufficient amounts were obtained for 12 CTR and 15 mCRS mice) in heparin-coated capillary tubes (Sarsted, Switzerland) and was centrifuged at 9400 × g at 4 °C for 4 min to obtain plasma. Basal plasma CORT levels were diluted 20 times and then were measured with an ELISA (Enzo Life Sciences, ADI-901-097 for corticosterone) according to the manufacturer’s protocol. Concentration values of CORT were calculated using a 4-parameter logistic fit.

Adrenal glands weight

Adrenal glands were dissected (due to loss during tissue collection, not available for one CTR mouse) and measured to determine adrenal gland weights normalized with body weights of each tested animal [weight adrenal gland in g/body weight in g) * 100].

Mitochondrial respiratory capacity

PFC and NAc from one hemisphere per animal, counterbalanced across samples [n = 16 CTR, n = 15 mCRS (due to loss during tissue collection, not available for one stressed animal) for PFC and n = 16/group for NAc], was used to measure mitochondrial respiration using the Oroboros Oxygraph 2 K (Oroboros Instruments, Innsbruck, Austria), as previously described⁸¹. Briefly, tissue was weight and homogenized in ice-cold respirometry medium (0.5 mM EGTA, 3 mM MgCl2, 60 mM potassium lactobionate, 20 mM taurine, 10 mM KH2PO4, 20 mM HEPES, 110 mM sucrose and 0.1% (w/v) BSA, pH = 7.1) with an eppendorf pestle. 2 mg of tissue was used to run a multisubstrate protocol at 37 °C to sequentially assess the various components of mitochondrial respiratory capacity through oxidative phosphorylation (OXPHOS): respiration due to complex I activity (Complex I) was measured by adding 5 mM ADP to a mixture of 2 mM malate, 10 mM pyruvate and 20 mM glutamate, followed by the addition of 10 mM succinate to subsequently stimulate complex II (Complex I + II). Respiration then was uncoupled to examine maximal capacitiy of the electron transport system (ETS) using the protonophore, carbonylcyanide 4 (trifluoromethoxy) phenylhydrazone (FCCP; successive titrations of 0.2 µM until maximal respiration rates were reached). The consumption in the uncoupled state due to complex II activity was determined by inhibiting complex I with the addition of 0.1 µM rotenone (ETS CII). In the last step, electron transport though complex III was inhibited to obtain the level of residual oxygen consumption (ROX) due to oxidating side reactions outside of mitochondrial respiration by adding 2 µM antimycin. The O₂ flux obtained in each step of the protocol was normalized by the wet weight of the tissue sample used for the analysis and corrected for ROX.

¹H-magnetic resonance spectroscopy (¹H-MRS)

Another batch of mice (n = 8/group) was devoted to in vivo ¹H-MRS measurements of brain metabolites, following procedures previously described in^18,54. Briefly, all experiments were performed on a horizontal 14.1 T/26 cm bore animal MR scanner (Magnex Scientific, Abingdon, UK) using a homemade quadrature ¹H-coil. The volume of interest (VOI) for spectroscopy was placed in the medial PFC (voxel size: 1.4 × 1.7 × 1.2 mm³) or in the bilateral NAc (voxel size: 4.1 × 1.4 × 1 mm³) after acquisition of a set of anatomical T₂-weighted images for localization. Field homogeneity was adjusted using FAST(EST)MAP to reach a typical water linewidth of 15 Hz in medial PFC and 20 Hz in NAc⁹⁸. Spectra were acquired with 20 blocks of 16 averages for medial PFC and 25 blocks of 16 averages for the NAc, leading to a scan duration of around 20 and 25 min respectively. After post-processing of the spectra, metabolite concentrations as well as the Cramér-Rao lower bounds (CRLB) were determined with LCModel using water as internal reference⁹⁹.

RNA isolation, cDNA synthesis, and quantitative RT-PCR

PFC and NAc from one hemisphere per animal, counterbalanced across samples, were snap frozen in liquid nitrogen until use for RNA extraction (n = 16/group). To this end, we used RNAqueous-Micro Total RNA Isolation Kit (ThermoFisher Scientific; #AM1931) and cDNA synthesis was done with the qScript cDNA SuperMix (QuantaBio; #733-1176). Quantitative RT-PCR (qRT-PCR) was performed using the SYBR Green PCR Master Mix (Applied Biosystems, Life Technologies, USA) at an ABI Prism 7900 Sequence Detection system (Applied Biosystems, Singapore) with eukaryotic elongation factor 1 (Ef1a) as a reference gene. Primer (Microsynth AG) sequences were as follows: Ef1a fw: tccacttggtcgctttgct, rv: cttcttgtccacagctttgatga; ND-1: fw: ggatgagcctcaaactccaa, rv: ggtcaggctggcagaagtaa; ND-3: fw: gcattctgactcccccaaat, rv: tgaattgctcatggtagtgga; ATP-6: fw: ccttccacaaggaactccaa, rv: ggtagctgttggtgggctaa. Gapdh: fw: tcaccaccatggagaaggc, rv: gctaagcagttggtggtgca.

Bulk RNA-sequencing

The extracted RNA was used for performing the transcriptomic analysis from PFC and NAc of mCRS and unstressed CTR mice [n = 15/group for PFC (two samples had to be removed from the analysis due to poor sequencing quality); n = 16/group for NAc] using bulk RNA (BRB)-sequencing¹⁰⁰. Briefly, each RNA sample was reverse transcribed in a 96-well plate using SuperScript II Reverse Transcriptase (Lifetech 18064014) with individual barcoded oligo-dT primers (Microsynth). Next, all the samples were pooled together, purified using the DNA Clean and Concentrator kit (Zymo Research #D4014), and treated with exonuclease I (NEB or New England BioLabs #M0293S). Double-stranded cDNA was generated by the second stand synthesis via the nick translation method. For that, a mix containing 2 μL of RNAse H (NEB, #M0297S), 1 μL of Escherichia coli DNA ligase (NEB, #M0205 L), 5 μL of E. coli DNA Polymerase (NEB, #M0209 L), 1 μL of dNTP (0 0.2 mM), 10 μL of 5 × Second Stand Buffer (100 mM Tris–HCl (pH 6.9, AppliChem, #A3452); 25 mM MgCl2 (Sigma, #M2670); 450 mM KCl (AppliChem, #A2939); 0.8 mM β-NAD Sigma, N1511); 60 mM (NH4)2SO4 (Fisher Scientific Acros, #AC20587); and 11 μL of water was added to 20 μL of ExoI-treated first-strand reaction on ice. The reaction was incubated at 16 °C for 2.5 h. Full-length double-stranded cDNA was purified with 30 μL (0.6 ×) of AMPure XP magnetic beads (Beckman Coulter, #A63881) and eluted in 20 μL of water. The Illumina compatible libraries were prepared by tagmentation of 5 ng of full-length double-stranded cDNA with 1 µL (11 μM) of in-house produced Tn5 enzyme. The final library was amplified 15 cycles and the fragments ranging 200–1000 bp were size-selected using AMPure beads (Beckman Coulter, #A63881) (first round 0.5 × beads, second 0.7 ×). The libraries were profiled with High Sensitivity NGS Fragment Analysis Kit (Advanced Analytical, #DNF-474) and measured with Qubit dsDNA HS Assay Kit (Invitrogen, #Q32851) prior to pooling and sequencing using the Illumina NextSeq 500 platform using a custom primer and the High Output v2 kit (75 cycles) (Illumina, #FC-404-2005). The library loading concentration was 2.2 pM and sequencing configuration as following: R1 6c/index 8 c/R2 70c.

Transcriptomic analysis

To assess whether changes in mitochondrial gene expression are a common response to chronic stress, we made use of a publicly available transcriptome data set [GEO accession number: GSE102556;²⁰]. To this end, we reanalyzed the raw reads of the mouse subseries of this study with a focus on mitochondrial associated gene expression by assessing differential gene expression in PFC (n = 19/group) and NAc (n = 20/group) between chronic variable stress and unstressed mice. To determine whether MDD subjects exhibit similar changes in mitochondrial gene signatures, we made use of a similar analysis pipeline and applied it on the subseries for MDD subjects of the aforementioned data set [GEO accession number: GSE102556;²⁰]. We assessed differential gene expression between MDD [NAc: n = 16, 8 males and 8 females; ventromedial (vm) PFC: n = 16, 7 males and 9 females] and healthy subjects [NAc: n = 17, 9 males and 8 females; vmPFC: n = 15, 9 males and 6 females] with available medication status.

Reads produced in this study or publicly available raw reads were mapped with STAR 2.4.0 g¹⁰¹ onto GRCm38/mm10 and GRCh38/hg38 genome assembly for mouse and human data, respectively. Uniquely mapped reads were counted for each gene locus using htseq 0.6.1¹⁰². We normalized count data by size factor and applied a variance stabilizing transformation for visualization purposes as suggested by Anders et al.¹⁰³. We utilized generalized linear model to assess differential gene expression with the help of the DESeq2 package¹⁰⁴. If not otherwise stated, we contrasted two groups for the mouse data [mCRS vs. CTR (data of this study), and CVS vs. CTR²⁰] using a Wald test. To statistically assess differential gene expression between human MDD subjects and healthy controls we used a likelihood-ratio to correct differential gene expression for effects from age, RNA quality (RIN), alcohol abuse, medication status and gender as suggested²⁰.

Transcription factor activity analysis

To predict transcription factor binding sites of transcription factors in promoters of differentially expressed genes, we employed MotEvo 1.03¹⁰⁵ and scanned ± 500 bp around the annotated transcription start site (TSS) of a gene. TSS definition were retrieved from¹⁰⁶ and weight matrices were retrieved from Swiss regulon¹⁰⁷. To compute transcription factor activity, we applied a penalized regression model described in Balwierz et al.¹⁰⁶.

Functional and gene set enrichment analysis

For gene set enrichment analysis we employed the camera function of the limma package¹⁰⁸ on a pre-ranked gene lists defined from the differential gene expression analysis. Gene sets were derived from several sources including AmiGO¹⁰⁹ for Gene Ontology Biological processes and KEGG¹¹⁰. The gene set of mitochondrial associated genes was retrieved from MitoCarta2.0¹¹¹.

Model selection

To assess whether a gene is better explained by the individual behavioral changes of an animal or the absence or presence of mCRS, we applied a model selection approach under the generalized linear model framework of DESeq2. To this end, we defined three models: (1) gene expression explained by z-score, (2) by presence or absence of mCRS or (3) no difference between animals. Each model was solved with generalized linear regression and Bayesian information criterion (BIC) was computed using the deviance of the fit to control for model complexity¹¹². We calculated the Schwarz weight to assess model confidence for each model. This analysis was done with animals for which the z-score could be determined and for which transcriptome data were available (n = 10 CTRs, n = 15 mCRS).

Statistics

Statistical analyses were performed with Prism 6.0 (Graphpad Software Inc.) and R (R Core Team, 2014). Unless otherwise stated, statistical significance was determined for the comparison of two groups by Student’s t-test for data following normal distribution or Mann–Whitney U-test for data following non-normal distribution. For analyses involving more than two groups, a 2-way analysis of variance (ANOVA) followed by Sidak’s post hoc test was applied. We log-transformed metabolite levels for spectroscopy and used a linear model to test for the contrast mCRS versus CTR. p values were corrected for multiple testing using the Benjamini–Hochberg method¹¹³. To test for differences between mCRS and CTR treated mice in respiration experiments, we constructed a linear mixed effects models using the R package lme4¹¹⁴. As the measurements for the mitochondrial respiration capacity experiments were performed in blocks of different days, we extracted the p values for the fixed effect treatment and controlled for day of measurement as a random effect. We reported the estimated marginal means of the model. All values represent the mean ± SEM. Asterisks indicate p value: *p < 0.05, **p < 0.01, ***p < 0.001.

Data availability

The generated datasets of this study were made publicly available at the GEO repository (GEO accession number: GSE148629) and can be accessed via the following link: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE148629.

References

McEwen, B. S. Physiology and neurobiology of stress and adaptation: Central role of the brain. Physiol. Rev. 87, 873–904 (2007).
Article PubMed Google Scholar
Sandi, C. Stress, cognitive impairment and cell adhesion molecules. Nat. Rev. Neurosci. 5, 917–930 (2004).
Article CAS PubMed Google Scholar
de Kloet, E. R., Joëls, M. & Holsboer, F. Stress and the brain: From adaptation to disease. Nat. Rev. Neurosci. 6, 463–475 (2005).
Article PubMed CAS Google Scholar
Kessler, R. C. et al. Lifetime prevalence and age-of-onset distributions of DSM-IV disorders in the National Comorbidity Survey Replication. Arch. Gen. Psychiatry 62, 593–602 (2005).
Article PubMed Google Scholar
Malhi, G. S. & Mann, J. J. Depression. Lancet (London, England) 392, 2299–2312 (2018).
Article Google Scholar
de Kloet, E. R. et al. Stress and depression: A crucial role of the mineralocorticoid receptor. J. Neuroendocrinol. 28 https://doi.org/10.1111/jne.12379 (2016).
Troubat, R. et al. Neuroinflammation and depression: A review. Eur. J. Neurosci. https://doi.org/10.1111/ejn.14720 (2020).
Article PubMed Google Scholar
Ward, J. et al. The genomic basis of mood instability: Identification of 46 loci in 363,705 UK Biobank participants, genetic correlation with psychiatric disorders, and association with gene expression and function. Mol. Psychiatry https://doi.org/10.1038/s41380-019-0439-8 (2019).
Article PubMed PubMed Central Google Scholar
Dudek, K. A. et al. Neurobiology of resilience in depression: Immune and vascular insights from human and animal studies. Eur. J. Neurosci. https://doi.org/10.1111/ejn.14547 (2019).
Article PubMed PubMed Central Google Scholar
Duman, R. S., Sanacora, G. & Krystal, J. H. Altered connectivity in depression: GABA and glutamate neurotransmitter deficits and reversal by novel treatments. Neuron 102, 75–90 (2019).
Article CAS PubMed PubMed Central Google Scholar
Muir, J., Lopez, J. & Bagot, R. C. Wiring the depressed brain: Optogenetic and chemogenetic circuit interrogation in animal models of depression. Neuropsychopharmacology 44, 1013–1026 (2019).
Article PubMed Google Scholar
Belleau, E. L., Treadway, M. T. & Pizzagalli, D. A. The impact of stress and major depressive disorder on hippocampal and medial prefrontal cortex morphology. Biol. Psychiatry 85, 443–453 (2019).
Article PubMed Google Scholar
Russo, S. J. & Nestler, E. J. The brain reward circuitry in mood disorders. Nat. Rev. Neurosci. 14, 609–625 (2013).
Article CAS PubMed Google Scholar
Treadway, M. T. & Pizzagalli, D. A. Imaging the pathophysiology of major depressive disorder—From localist models to circuit-based analysis. Biol. Mood Anxiety Disord. 4, 5 (2014).
Article PubMed PubMed Central CAS Google Scholar
Kennedy, S. H. et al. Changes in regional brain glucose metabolism measured with positron emission tomography after paroxetine treatment of major depression. Am. J. Psychiatry 158, 899–905 (2001).
Article CAS PubMed Google Scholar
Sequeira, A. et al. Patterns of gene expression in the limbic system of suicides with and without major depression. Mol. Psychiatry 12, 640–655 (2007).
Article CAS PubMed Google Scholar
McKlveen, J. M. et al. Chronic stress increases prefrontal inhibition: A mechanism for stress-induced prefrontal dysfunction. Biol. Psychiatry 80, 754–764 (2016).
Article PubMed PubMed Central Google Scholar
Larrieu, T. et al. Hierarchical status predicts behavioral vulnerability and nucleus accumbens metabolic profile following chronic social defeat stress. Curr. Biol. 27, 2202-2210.e4 (2017).
Article CAS PubMed Google Scholar
Chandra, R. et al. Reduced Slc6a15 in nucleus Accumbens D2-neurons underlies stress susceptibility. J. Neurosci. 37, 6527–6538 (2017).
Article CAS PubMed PubMed Central Google Scholar
Labonté, B. et al. Sex-specific transcriptional signatures in human depression. Nat. Med. 23, 1102–1111 (2017).
Article PubMed PubMed Central CAS Google Scholar
Seney, M. L. et al. Opposite molecular signatures of depression in men and women. Biol. Psychiatry 84, 18–27 (2018).
Article CAS PubMed PubMed Central Google Scholar
Nagy, C. et al. Single-nucleus transcriptomics of the prefrontal cortex in major depressive disorder implicates oligodendrocyte precursor cells and excitatory neurons. Nat. Neurosci. https://doi.org/10.1038/s41593-020-0621-y (2020).
Article PubMed Google Scholar
Krishnan, V. et al. Molecular adaptations underlying susceptibility and resistance to social defeat in brain reward regions. Cell 131, 391–404 (2007).
Article CAS PubMed Google Scholar
Bagot, R. C. et al. Circuit-wide transcriptional profiling reveals brain region-specific gene networks regulating depression susceptibility. Neuron 90, 969–983 (2016).
Article CAS PubMed PubMed Central Google Scholar
Hodes, G. E. et al. Sex Differences in nucleus accumbens transcriptome profiles associated with susceptibility versus resilience to subchronic variable stress. J. Neurosci. 35, 16362–16376 (2015).
Article CAS PubMed PubMed Central Google Scholar
Scarpa, J. R. et al. Shared transcriptional signatures in major depressive disorder and mouse chronic stress models. Biol. Psychiatry https://doi.org/10.1016/j.biopsych.2019.12.029 (2020).
Article PubMed PubMed Central Google Scholar
Picard, M. et al. Mitochondrial functions modulate neuroendocrine, metabolic, inflammatory, and transcriptional responses to acute psychological stress. Proc. Natl. Acad. Sci. 112, E6614–E6623 (2015).
Article CAS PubMed PubMed Central Google Scholar
Picard, M., McEwen, B. S., Epel, E. S. & Sandi, C. An energetic view of stress: Focus on mitochondria. Front. Neuroendocrinol. 49, 72–85 (2018).
Article CAS PubMed PubMed Central Google Scholar
Filiou, M. D. & Sandi, C. Anxiety and brain mitochondria: A bidirectional crosstalk. Trends Neurosci. 42, 573–588 (2019).
Article CAS PubMed Google Scholar
Morava, E. & Kozicz, T. Mitochondria and the economy of stress (mal)adaptation. Neurosci. Biobehav. Rev. 37, 668–680 (2013).
Article CAS PubMed Google Scholar
Petschner, P. et al. Genes linking mitochondrial function, cognitive impairment and depression are associated with endophenotypes serving precision medicine. Neuroscience 370, 207–217 (2018).
Article CAS PubMed Google Scholar
Kasahara, T. & Kato, T. What can mitochondrial dna analysis tell us about mood disorders?. Biol. Psychiatry 83, 731–738 (2018).
Article CAS PubMed Google Scholar
Torrell, H. et al. Mitochondrial DNA (mtDNA) in brain samples from patients with major psychiatric disorders: Gene expression profiles, mtDNA content and presence of the mtDNA common deletion. Am. J. Med. Genet. B. Neuropsychiatr. Genet. 162B, 213–223 (2013).
Article PubMed CAS Google Scholar
Wang, Q. & Dwivedi, Y. Transcriptional profiling of mitochondria associated genes in prefrontal cortex of subjects with major depressive disorder. World J. Biol. Psychiatry 18, 592–603 (2017).
Article PubMed Google Scholar
Misiewicz, Z. et al. Multi-omics analysis identifies mitochondrial pathways associated with anxiety-related behavior. PLoS Genet. 15, e1008358 (2019).
Article PubMed PubMed Central CAS Google Scholar
Duarte, J. M. N., Lei, H., Mlynárik, V. & Gruetter, R. The neurochemical profile quantified by in vivo ¹H NMR spectroscopy. Neuroimage 61, 342–362 (2012).
Article CAS PubMed Google Scholar
Münch, C. & Harper, J. W. Mitochondrial unfolded protein response controls matrix pre-RNA processing and translation. Nature 534, 710–713 (2016).
Article ADS PubMed PubMed Central CAS Google Scholar
Fiorese, C. J. et al. The transcription factor ATF5 mediates a mammalian mitochondrial UPR. Curr. Biol. 26, 2037–2043 (2016).
Article CAS PubMed PubMed Central Google Scholar
Quirós, P. M. et al. Multi-omics analysis identifies ATF4 as a key regulator of the mitochondrial stress response in mammals. J. Cell Biol. 216, 2027–2045 (2017).
Article PubMed PubMed Central CAS Google Scholar
Melber, A. & Haynes, C. M. UPRmt regulation and output: A stress response mediated by mitochondrial-nuclear communication. Cell Res. 28, 281–295 (2018).
Article CAS PubMed PubMed Central Google Scholar
Polman, J. A. E. et al. Glucocorticoids modulate the mTOR pathway in the hippocampus: Differential effects depending on stress history. Endocrinology 153, 4317–4327 (2012).
Article CAS PubMed Google Scholar
Nasca, C., Bigio, B., Zelli, D., Nicoletti, F. & McEwen, B. S. Mind the gap: Glucocorticoids modulate hippocampal glutamate tone underlying individual differences in stress susceptibility. Mol. Psychiatry 20, 755–763 (2015).
Article CAS PubMed Google Scholar
Kupferberg, A., Bicks, L. & Hasler, G. Social functioning in major depressive disorder. Neurosci. Biobehav. Rev. 69, 313–332 (2016).
Article PubMed Google Scholar
Salamone, J. D., Yohn, S. E., López-Cruz, L., San Miguel, N. & Correa, M. Activational and effort-related aspects of motivation: Neural mechanisms and implications for psychopathology. Brain 139, 1325–1347 (2016).
Article PubMed PubMed Central Google Scholar
Grahek, I., Shenhav, A., Musslick, S., Krebs, R. M. & Koster, E. H. W. Motivation and cognitive control in depression. Neurosci. Biobehav. Rev. 102, 371–381 (2019).
Article PubMed PubMed Central Google Scholar
Hamilton, P. J. et al. Multi-OMIC analysis of brain and serum from chronically-stressed mice reveals network disruptions in purine metabolism, fatty acid beta-oxidation, and antioxidant activity that are reversed by antidepressant treatment. bioRxiv https://doi.org/10.1101/490748 (2018).
Article Google Scholar
Holmes, A. & Wellman, C. L. Stress-induced prefrontal reorganization and executive dysfunction in rodents. Neurosci. Biobehav. Rev. 33, 773–783 (2009).
Article PubMed Google Scholar
Shansky, R. M. & Morrison, J. H. Stress-induced dendritic remodeling in the medial prefrontal cortex: Effects of circuit, hormones and rest. Brain Res. 1293, 108–113 (2009).
Article CAS PubMed PubMed Central Google Scholar
Duman, C. H. & Duman, R. S. Spine synapse remodeling in the pathophysiology and treatment of depression. Neurosci. Lett. 601, 20–29 (2015).
Article CAS PubMed PubMed Central Google Scholar
Thompson, S. M. et al. An excitatory synapse hypothesis of depression. Trends Neurosci. 38, 279–294 (2015).
Article CAS PubMed PubMed Central Google Scholar
Khibnik, L. A. et al. Stress and cocaine trigger divergent and cell type-specific regulation of synaptic transmission at single spines in nucleus accumbens. Biol. Psychiatry 79, 898–905 (2016).
Article CAS PubMed Google Scholar
Francis, T. C. et al. Molecular basis of dendritic atrophy and activity in stress susceptibility. Mol. Psychiatry 22, 1512–1519 (2017).
Article CAS PubMed PubMed Central Google Scholar
Gilabert-Juan, J., Castillo-Gomez, E., Guirado, R., Moltó, M. D. & Nacher, J. Chronic stress alters inhibitory networks in the medial prefrontal cortex of adult mice. Brain Struct. Funct. 218, 1591–1605 (2013).
Article CAS PubMed Google Scholar
Cherix, A. et al. Metabolic signature in nucleus accumbens for anti-depressant-like effects of acetyl-L-carnitine. Elife 9, e50631 (2020).
Article CAS PubMed PubMed Central Google Scholar
Lopes, S. et al. Tau deletion prevents stress-induced dendritic atrophy in prefrontal cortex: Role of synaptic mitochondria. Cereb. Cortex https://doi.org/10.1093/cercor/bhw057 (2016).
Article Google Scholar
Hunter, R. G. et al. Stress and corticosteroids regulate rat hippocampal mitochondrial DNA gene expression via the glucocorticoid receptor. Proc. Natl. Acad. Sci. USA 113, 9099–9104 (2016).
Article CAS PubMed PubMed Central Google Scholar
Emmerzaal, T. L. et al. Impaired mitochondrial complex I function as a candidate driver in the biological stress response and a concomitant stress-induced brain metabolic reprogramming in male mice. Transl. Psychiatry 10, 176 (2020).
Article CAS PubMed PubMed Central Google Scholar
Grigoruţă, M., Martínez-Martínez, A., Dagda, R. Y. & Dagda, R. K. Psychological stress phenocopies brain mitochondrial dysfunction and motor deficits as observed in a parkinsonian rat model. Mol. Neurobiol. 57, 1781–1798 (2020).
Article PubMed CAS Google Scholar
Kuffner, K. et al. Major depressive disorder is associated with impaired mitochondrial function in skin fibroblasts. Cells 9, 884 (2020).
Article PubMed Central Google Scholar
Drevets, W. C., Ongür, D. & Price, J. L. Reduced glucose metabolism in the subgenual prefrontal cortex in unipolar depression. Mol. Psychiatry 3, 190–191 (1998).
Article CAS PubMed Google Scholar
Drevets, W. C., Price, J. L. & Furey, M. L. Brain structural and functional abnormalities in mood disorders: Implications for neurocircuitry models of depression. Brain Struct. Funct. 213, 93–118 (2008).
Article PubMed PubMed Central Google Scholar
MacDonald, K. et al. Biomarkers for major depressive and bipolar disorders using metabolomics: A systematic review. Am. J. Med. Genet. B. Neuropsychiatr. Genet. 180, 122–137 (2019).
Article PubMed Google Scholar
van der Kooij, M. A. et al. Chronic social stress-induced hyperglycemia in mice couples individual stress susceptibility to impaired spatial memory. Proc. Natl. Acad. Sci. USA 115, E10187–E10196 (2018).
Article PubMed CAS PubMed Central Google Scholar
Mergenthaler, P., Lindauer, U., Dienel, G. A. & Meisel, A. Sugar for the brain: The role of glucose in physiological and pathological brain function. Trends Neurosci. 36, 587–597 (2013).
Article CAS PubMed PubMed Central Google Scholar
Hall, C. N., Klein-Flugge, M. C., Howarth, C. & Attwell, D. Oxidative phosphorylation, not glycolysis, powers presynaptic and postsynaptic mechanisms underlying brain information processing. J. Neurosci. 32, 8940–8951 (2012).
Article CAS PubMed PubMed Central Google Scholar
McEwen, B. S. Allostasis and the epigenetics of brain and body health over the life course. JAMA Psychiatry 74, 551 (2017).
Article PubMed Google Scholar
Picard, M., Juster, R.-P. & McEwen, B. S. Mitochondrial allostatic load puts the ‘gluc’ back in glucocorticoids. Nat. Rev. Endocrinol. 10, 303–310 (2014).
Article CAS PubMed Google Scholar
Du, J. et al. Dynamic regulation of mitochondrial function by glucocorticoids. Proc. Natl. Acad. Sci. 106, 3543–3548 (2009).
Article ADS CAS PubMed PubMed Central Google Scholar
Musazzi, L. et al. Acute inescapable stress rapidly increases synaptic energy metabolism in prefrontal cortex and alters working memory performance. Cereb. Cortex 29, 4948–4957 (2019).
Article PubMed Google Scholar
McEwen, B. S. Protection and damage from acute and chronic stress: Allostasis and allostatic overload and relevance to the pathophysiology of psychiatric disorders. Ann. N. Y. Acad. Sci. 1032, 1–7 (2004).
Article ADS PubMed Google Scholar
Pacelli, C. et al. Elevated mitochondrial bioenergetics and axonal arborization size are key contributors to the vulnerability of dopamine neurons. Curr. Biol. 25, 2349–2360 (2015).
Article CAS PubMed Google Scholar
Surmeier, D. J. Calcium, ageing, and neuronal vulnerability in Parkinson’s disease. Lancet Neurol. 6, 933–938 (2007).
Article CAS PubMed Google Scholar
Chinta, S. J. & Andersen, J. K. Redox imbalance in Parkinson’s disease. Biochim. Biophys. Acta Gen. Subj. 1780, 1362–1367 (2008).
Article CAS Google Scholar
Licznerski, P. & Duman, R. S. Remodeling of axo-spinous synapses in the pathophysiology and treatment of depression. Neuroscience 251, 33–50 (2013).
Article CAS PubMed Google Scholar
McEwen, B. S., Nasca, C. & Gray, J. D. Stress effects on neuronal structure: Hippocampus, amygdala, and prefrontal cortex. Neuropsychopharmacology 41, 3–23 (2016).
Article CAS PubMed Google Scholar
Stewart, M. G. et al. Stress suppresses and learning induces plasticity in CA3 of rat hippocampus: A three-dimensional ultrastructural study of thorny excrescences and their postsynaptic densities. Neuroscience 131, 43–54 (2005).
Article CAS PubMed Google Scholar
Magariños, A. M. & McEwen, B. S. Stress-induced atrophy of apical dendrites of hippocampal CA3c neurons: Comparison of stressors. Neuroscience 69, 83–88 (1995).
Article PubMed Google Scholar
Vyas, A., Pillai, A. G. & Chattarji, S. Recovery after chronic stress fails to reverse amygdaloid neuronal hypertrophy and enhanced anxiety-like behavior. Neuroscience 128, 667–673 (2004).
Article CAS PubMed Google Scholar
Zhang, J.-Y. et al. Chronic stress remodels synapses in an amygdala circuit-specific manner. Biol. Psychiatry 85, 189–201 (2019).
Article CAS PubMed Google Scholar
Liston, C. et al. Stress-induced alterations in prefrontal cortical dendritic morphology predict selective impairments in perceptual attentional set-shifting. J. Neurosci. 26, 7870–7874 (2006).
Article CAS PubMed PubMed Central Google Scholar
Hollis, F. et al. Mitochondrial function in the brain links anxiety with social subordination. Proc. Natl. Acad. Sci. USA 112, 15486–15491 (2015).
Article ADS CAS PubMed PubMed Central Google Scholar
Timmer, M. & Sandi, C. A role for glucocorticoids in the long-term establishment of a social hierarchy. Psychoneuroendocrinology 35, 1543–1552 (2010).
Article CAS PubMed Google Scholar
Weger, M. et al. Increased brain glucocorticoid actions following social defeat in rats facilitates the long-term establishment of social subordination. Physiol. Behav. 186, 31–36 (2018).
Article CAS PubMed Google Scholar
Larrieu, T. & Sandi, C. Stress-induced depression: Is social rank a predictive risk factor?. BioEssays 40, e1800012 (2018).
Article PubMed Google Scholar
Barik, J. et al. Chronic stress triggers social aversion via glucocorticoid receptor in dopaminoceptive neurons. Science 339, 332–335 (2013).
Article ADS CAS PubMed Google Scholar
Douma, E. H. & de Kloet, E. R. Stress-induced plasticity and functioning of ventral tegmental dopamine neurons. Neurosci. Biobehav. Rev. 108, 48–77 (2020).
Article CAS PubMed Google Scholar
Adzic, M. et al. Brain region- and sex-specific modulation of mitochondrial glucocorticoid receptor phosphorylation in fluoxetine treated stressed rats: Effects on energy metabolism. Psychoneuroendocrinology 38, 2914–2924 (2013).
Article CAS PubMed Google Scholar
Moutsatsou, P. et al. Localization of the glucocorticoid receptor in rat brain mitochondria. Arch. Biochem. Biophys. 386, 69–78 (2001).
Article CAS PubMed Google Scholar
Psarra, A.-M.G. & Sekeris, C. E. Glucocorticoids induce mitochondrial gene transcription in HepG2 cells: Role of the mitochondrial glucocorticoid receptor. Biochim. Biophys. Acta 1813, 1814–1821 (2011).
Article CAS PubMed Google Scholar
Nasca, C. et al. L-acetylcarnitine causes rapid antidepressant effects through the epigenetic induction of mGlu2 receptors. Proc. Natl. Acad. Sci. USA 110, 4804–4809 (2013).
Article ADS CAS PubMed PubMed Central Google Scholar
Poirier, G. L., Imamura, N., Zanoletti, O. & Sandi, C. Social deficits induced by peripubertal stress in rats are reversed by resveratrol. J. Psychiatr. Res. 57, 157–164 (2014).
Article PubMed Google Scholar
Hollis, F., Mitchell, E. S., Canto, C., Wang, D. & Sandi, C. Medium chain triglyceride diet reduces anxiety-like behaviors and enhances social competitiveness in rats. Neuropharmacology 138, 245–256 (2018).
Article CAS PubMed Google Scholar
Weger, M. & Sandi, C. High anxiety trait: A vulnerable phenotype for stress-induced depression. Neurosci. Biobehav. Rev. 87, 27–37 (2018).
Article PubMed Google Scholar
Golden, S. A., Covington, H. E., Berton, O. & Russo, S. J. A standardized protocol for repeated social defeat stress in mice. Nat. Protoc. 6, 1183–1191 (2011).
Article CAS PubMed PubMed Central Google Scholar
Anacker, C. et al. Neuroanatomic differences associated with stress susceptibility and resilience. Biol. Psychiatry 79, 840–849 (2016).
Article PubMed Google Scholar
Porsolt, R. D., Bertin, A. & Jalfre, M. Behavioral despair in mice: A primary screening test for antidepressants. Arch. Int. Pharmacodyn. Ther. 229, 327–336 (1977).
CAS PubMed Google Scholar
Guilloux, J.-P., Seney, M., Edgar, N. & Sibille, E. Integrated behavioral z-scoring increases the sensitivity and reliability of behavioral phenotyping in mice: Relevance to emotionality and sex. J. Neurosci. Methods 197, 21–31 (2011).
Article PubMed PubMed Central Google Scholar
Gruetter, R. & Tkác, I. Field mapping without reference scan using asymmetric echo-planar techniques. Magn. Reson. Med. 43, 319–323 (2000).
Article CAS PubMed Google Scholar
Provencher, S. W. Automatic quantitation of localized in vivo ¹H spectra with LCModel. NMR Biomed. 14, 260–264 (2001).
Article CAS PubMed Google Scholar
Alpern, D. et al. BRB-seq: Ultra-affordable high-throughput transcriptomics enabled by bulk RNA barcoding and sequencing. Genome Biol. 20, 71 (2019).
Article PubMed PubMed Central Google Scholar
Dobin, A. et al. STAR: Ultrafast universal RNA-seq aligner. Bioinformatics 29, 15–21 (2013).
Article CAS PubMed Google Scholar
Anders, S., Pyl, P. T. & Huber, W. HTSeq—A Python framework to work with high-throughput sequencing data. Bioinformatics 31, 166–169 (2015).
Article CAS PubMed Google Scholar
Anders, S. & Huber, W. Differential expression analysis for sequence count data. Genome Biol. 11, R106 (2010).
Article CAS PubMed PubMed Central Google Scholar
Love, M. I., Huber, W. & Anders, S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 15, 550 (2014).
Article PubMed PubMed Central CAS Google Scholar
Arnold, P., Erb, I., Pachkov, M., Molina, N. & van Nimwegen, E. MotEvo: Integrated Bayesian probabilistic methods for inferring regulatory sites and motifs on multiple alignments of DNA sequences. Bioinformatics 28, 487–494 (2012).
Article CAS PubMed Google Scholar
Balwierz, P. J. et al. Methods for analyzing deep sequencing expression data: Constructing the human and mouse promoterome with deepCAGE data. Genome Biol. 10, R79 (2009).
Article PubMed PubMed Central CAS Google Scholar
Pachkov, M., Balwierz, P. J., Arnold, P., Ozonov, E. & van Nimwegen, E. SwissRegulon, a database of genome-wide annotations of regulatory sites: Recent updates. Nucleic Acids Res. 41, D214–D220 (2013).
Article CAS PubMed Google Scholar
Ritchie, M. E. et al. Limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res. 43, e47 (2015).
Article PubMed PubMed Central CAS Google Scholar
Carbon, S. et al. AmiGO: Online access to ontology and annotation data. Bioinformatics 25, 288–289 (2009).
Article CAS PubMed Google Scholar
Kanehisa, M. & Goto, S. KEGG: Kyoto encyclopedia of genes and genomes. Nucleic Acids Res. 28, 27–30 (2000).
Article CAS PubMed PubMed Central Google Scholar
Calvo, S. E., Clauser, K. R. & Mootha, V. K. MitoCarta2.0: An updated inventory of mammalian mitochondrial proteins. Nucleic Acids Res. 44, D1251–D1257 (2016).
Article CAS PubMed Google Scholar
Kass, R. E. & Raftery, A. E. Bayes factors. J. Am. Stat. Assoc. 90, 773–795 (1995).
Article MathSciNet MATH Google Scholar
Benjamini, Y. & Hochberg, Y. Controlling the false discovery rate: A practical and powerful approach to multiple testing. J. R. Stat. Soc. Ser. B 57, 289–300 (1995).
MathSciNet MATH Google Scholar
Bates, D., Mächler, M., Bolker, B. & Walker, S. Fitting linear mixed-effects models using lme4. J. Stat. Softw. 67, 1–48 (2015).
Article Google Scholar

Download references

Acknowledgements

We would like to thank Dr Benjamin D. Weger for scientific discussions and critically reading the manuscript, and Dr Laia Morato Fornaguera and Dr Thomas Larrieu for scientific discussions. We acknowledge Olivia Zanoletti and Isabelle Guillot de Suduiraut for excellent technical assistance. This project has been supported by grants from the Swiss National Science Foundation [176206; NCCR Synapsy: 51NF40-158776 and -185897] and intramural funding from the EPFL (to CS). MW was awarded with a Marie Curie Intra-European Fellowships for Career Development (EU Project 748051, MitoDyaD).

Author information

Antoine Cherix
Present address: Nuffield Department of Clinical Neurosciences, John Radcliffe Hospital, University of Oxford, Oxford, OX3 9DU, England, UK

Authors and Affiliations

Laboratory of Behavioral Genetics, Brain Mind Institute, École Polytechnique Fédérale de Lausanne, 1015, Lausanne, Switzerland
Meltem Weger, Sriparna Ghosal, Jocelyn Grosse & Carmen Sandi
Laboratory of Systems Biology and Genetics, Institute of Bioengineering, École Polytechnique Fédérale de Lausanne, 1015, Lausanne, Switzerland
Daniel Alpern, Julie Russeil & Bart Deplancke
Laboratory for Functional and Metabolic Imaging, École Polytechnique Fédérale de Lausanne, 1015, Lausanne, Switzerland
Antoine Cherix & Rolf Gruetter
Departement of Endocrinology and Metabolic Disease, Leiden University Medical Center, Leiden, The Netherlands
E. Ronald de Kloet
Institute for Molecular Bioscience, The University of Queensland, St. Lucia, QLD, 4072, Australia
Meltem Weger
Swiss Institute of Bioinformatics, 1015, Lausanne, Switzerland
Daniel Alpern & Bart Deplancke

Authors

Meltem Weger
View author publications
You can also search for this author in PubMed Google Scholar
Daniel Alpern
View author publications
You can also search for this author in PubMed Google Scholar
Antoine Cherix
View author publications
You can also search for this author in PubMed Google Scholar
Sriparna Ghosal
View author publications
You can also search for this author in PubMed Google Scholar
Jocelyn Grosse
View author publications
You can also search for this author in PubMed Google Scholar
Julie Russeil
View author publications
You can also search for this author in PubMed Google Scholar
Rolf Gruetter
View author publications
You can also search for this author in PubMed Google Scholar
E. Ronald de Kloet
View author publications
You can also search for this author in PubMed Google Scholar
Bart Deplancke
View author publications
You can also search for this author in PubMed Google Scholar
Carmen Sandi
View author publications
You can also search for this author in PubMed Google Scholar

Contributions

M.W. and C.S. designed the research. M.W., D.A., A.C., S.G., J.G., and J.M.R. performed the research. M.W., D.A., and A.C. analyzed the data. M.W., B.D., R.G., and C.S. provided funding and/or supervision. M.W. wrote the first draft and C.S. and E.R.K. contributed to the data interpretation and revision of the manuscript. All authors reviewed and approved the manuscript.

Corresponding author

Correspondence to Carmen Sandi.

Ethics declarations

Competing interests

The authors declare no competing interests.

Additional information

Publisher's note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary Information.

Supplementary Table 1.

Supplementary Table 2.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions

About this article

Cite this article

Weger, M., Alpern, D., Cherix, A. et al. Mitochondrial gene signature in the prefrontal cortex for differential susceptibility to chronic stress. Sci Rep 10, 18308 (2020). https://doi.org/10.1038/s41598-020-75326-9

Download citation

Received: 26 June 2020
Accepted: 14 October 2020
Published: 27 October 2020
DOI: https://doi.org/10.1038/s41598-020-75326-9

This article is cited by

Brain mitochondrial diversity and network organization predict anxiety-like behavior in male mice
- Ayelet M. Rosenberg
- Manish Saggar
- Martin Picard
Nature Communications (2023)
The cortisol switch between vulnerability and resilience
- E. Ronald de Kloet
- Marian Joëls
Molecular Psychiatry (2023)
Transcriptomic analysis reveals mitochondrial pathways associated with distinct adolescent behavioral phenotypes and stress response
- Thamyris Santos-Silva
- Doğukan Hazar Ülgen
- Felipe V. Gomes
Translational Psychiatry (2023)

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.