Brain mitochondria in behavior – more than a powerhouse

Abstract

Rosenberg and colleagues developed a miniaturized assay of mitochondrial content and energy transformation capacity that allowed them to assess mitochondrial features and activities across the brain in male mice. Their findings provide a comprehensive brain-wide map of mitochondrial associations with stress-induced behaviors and highlight mitochondria as dynamic, spatially-organized organelles.

Mitochondria are dynamic organelles found in nearly all mammalian cells except for red blood cells. Mitochondria function in a number of processes that are crucial for cell survival and health, including the vital supply of energy in the form of ATP. Hence, historically, mitochondria have been referred to as “the powerhouse of the cell” and their importance confined to energy production. However, it is becoming increasingly clear that mitochondria are more than just a powerhouse. Mitochondrial respiration in the nucleus accumbens and prefrontal cortex exhibits individual differences linked to trait anxiety and pharmacological manipulation of respiration directly modulates social competitive behavior [1,2]. Mitochondria also mediate the physiological response to stress. Manipulation of mitochondrial gene expression induced respiratory chain deficits and altered redox balance and energy exchange that perturbed neuroendocrine, metabolic, transcriptional, and inflammatory responses to an acute stress challenge, demonstrating a pivotal relationship between mitochondrial respiratory and redox capacities and the hypothalamic pituitary adrenal (HPA) axis [3,4]. However, studies have been limited to single brain regions, and the underlying mechanism linking brain mitochondrial function to behavior has so far been unclear. Rosenberg and colleagues developed a high throughput assay and analysis pipeline that allowed them to examine brain-wide mitochondrial responses in association with behaviors to identify networks associated with stress-induced behaviors [5].

Miniaturized assays to robustly assess mitochondrial health

The gold standard for assessing mitochondrial electron transport and oxidative phosphorylation is traditionally respirometry [6], which has constrained sample sizes and tissues, due to low throughput and minimum volume requirements, respectively. While some studies include multiple quantifications of mitochondrial measures, not all do, and few integrate these assays into a meaningful biological assessment. In their recent publication in Nature Communications, Rosenberg and colleagues modified a high-throughput assay developed from previous work in immune cells [7] to create a miniaturized high-throughput assay for sub-milligram quantification of activity measures of mitochondrial respiratory complexes I, II, and IV, mitochondrial DNA (mtDNA) density, and citrate synthase integrated into a single index of mitochondrial health (Mitochondrial Health Index, MHI) [5]. With this assay in hand, the authors were able to simultaneously assess 17 brain regions from a single male mouse to generate a systematic map of brain-region specific mitochondrial health that incorporates both mitochondrial abundance and energy production capacity under either naïve conditions, following chronic corticosterone treatment, or after chronic social defeat stress [5]. Each animal was assessed for known behavioral phenotypes of stress exposure – increased anxiety-like behavior, hyponeophagia, and approach avoidance using the social interaction, novelty suppressed feeding, elevated plus maze, and open field tests [5]. This approach generated a spectrum of diverse behavior that allowed for identification of associations with mitochondrial phenotypes using the MHI [5].

Behaviorally relevant and distinct mitochondrial phenotypes exist across the male mouse brain

Previous work identified a role for mitochondrial respiratory capacity in the direct modulation of behavior and hinted at the tantalizing possibility of brain region-specific relationships with behavior [1,2,8]. As analyses were confined to a small subset of two-three regions [1,2,8,9], the existence of specialized mitochondrial phenotypes in the brain was unclear. Here, the breadth of analyses across the brain is a true strength of this study, allowing for identification of potential “mitotypes” and mitochondrial networks associated with behavior. Even within an inbred mouse strain (C57Bl/6J), analyses revealed a surprising amount of variation in mitochondrial features both between naïve individuals and across sampled brain regions [5]. This suggests brain region-specific specialization of mitochondrial energy transformation capacity for regional needs, which opens the door to numerous follow up questions. For instance, how do individual differences in mitochondrial energy capacity arise, and might these differences underlie individual susceptibilities to pathology? Significant associations between MHI and anxiety-like behavior (elevated plus maze and open field) and social avoidance behavior [5] support previous suggestions of specificity between anxiety-like behavior and corticostriatal mitochondrial measures [1,2,8,9]. Interestingly, exposure to stressors induced, on average, opposing directional effects on MHI according to stressor-type (corticosterone vs chronic social defeat stress) [5]. Measures of stress effects on mitochondrial energy capacity have been limited to targeted brain regions [9] or peripheral measures in leukocytes [10]. Here, chronic social defeat stress induced mitochondrial responses throughout the brain, while chronic corticosterone treatment exerted effects in more specific, targeted regions [5]. The identification of distinct mitochondrial networks suggests that individuals might recruit different networks to respond to specific challenges according to energetic needs. Importantly, a small subset of animals was allowed to recover from chronic social defeat revealing mitochondrial measures only marginally different from naïve individuals [5]. This important finding, though statistically underpowered, emphasizes the dynamic and adaptable nature of mitochondria to environmental stimuli and further highlights the need for future studies of temporal dynamics surrounding the effects of stress on brain bioenergetics. Interestingly, mitochondrial assessments of peripheral tissues such as heart, liver, muscle, and adrenal gland exhibited less variability and notably did not correlate with brain MHI [5]. This reveals a distinct disconnect between brain and peripheral MHI that may reflect a specificity of MHI to brain. Further studies examining the relationship between peripheral and central mitochondrial respiratory capacity will be crucial in determining the level of independence and whether peripheral mitochondrial energy capacity can serve as indicators for brain mitochondrial energy capacity. Notably absent from this study was assessment of MHI in circulating immune cells, which may better reflect brain mitochondrial states. Moreover, the MHI captures maximal activity levels from frozen samples, thus functional measures of mitochondrial respiratory complexes (i.e., respirometry in fresh tissue) may reveal unidentified relationships.

Rosenberg and colleagues have produced a landmark study connecting mitochondrial health across the brain with behavior. This study provides a wealth of evidence for mitochondria as dynamic organelles with active roles in modulating behaviors and perhaps individual differences in stress susceptibility. The importance of mitochondria to health and medicine is increasing and while this study fundamentally extends our knowledge of brain mitochondrial biology, there are still a number of open questions (Figure 1). Stress has different effects depending on the life stage of exposure, with distinct patterns between males and females [11]. To examine the broad number of brain regions and manipulations, Rosenberg and colleagues necessarily focused on a single sex and age in an inbred strain – middle-aged male C57Bl/6J mice. Future studies determining mitochondrial networks across the brain in females and both sexes at earlier and/or different developmental time points in outbred strains or other species will be vital in understanding the stability, diversity, and evolutionary conservation of brain mitochondrial networks. Notably, similar association studies in the human brain are currently lacking. While pre- and post-mortem factor effects on mitochondria may limit direct translation, the miniaturized assay presented here may facilitate the establishment of mitochondrial networks in post-mortem human brain tissues, allowing for future translational studies. The mechanism underlying precisely how mitochondrial networks modulate behavior is still unclear. Mitochondria have a variety of functions that support neuronal health, including calcium buffering, signaling, synapse remodeling, and of course, ATP production [12]. Developing a spatially resolved understanding of brain mitochondrial biology will help to resolve the energetic constraints on brain function and behavior. Future studies isolating how stress alters mitochondrial features and capacities to modulate behavior as well as how individual brain regions recruit different mitochondrial networks will be crucial in advancing our understanding and discovery of new therapies that target mitochondria.

Figure 1. Experimental overview with open questions.

Animals were divided into four groups; naïve, corticosterone treatment (CORT), chronic social defeat stress (CSDS), CSDS + recovery. Following exposure, animals were selectively placed in behavioral tests of anxiety-like behavior (open field (OF); elevated plus maze (EPM)), approach-avoidance (social interaction (SI)), and hyponeophagia (novelty suppressed feeding (NSF)). Brain regions were selected based on associations with anxiety-related, social, cognitive, and mitochondrial disorders to be compared with previous behavioral results. Tissue samples were analyzed via spectrophotometry to assess enzymatic activity of respiratory chain (RC) complexes I, II, and IV, citrate synthase, mitochondrial density, and DNA copy number. Multi-slice community detection analysis showed three distinct networks of brain regions defined as 1) cortico-striatal network; 2) salience and spatial navigation network; and 3) threat response network which correlated with the observed behavioral outputs. Further studies are needed to examine the effects of stress at different developmental time points, in female brain mitochondria, in different species, and to understand the mechanisms by which mitochondrial networks modulate behavior. Finally, establishing whether similar mitochondrial networks are present in the human brain will facilitate future translational studies linking mitochondrial networks to behavior. This figure was created using BioRender (https://biorender.com/).