Background

Oxidative damage is associated with the aetiology of many diseases, including neurodegenerative disorders given that the brain is an exceptionally vulnerable tissue to oxidative stress as well as to age-related alterations (Cobley et al., 2018; Mattson and Arumugam, 2018). Yet there are still important gaps in the understanding of reactive oxygen species (ROS) pathophysiology in the brain. The deleterious effects associated with ROS in situations of redox stress are in contrast to the increasing evidence suggesting that physiological processes are also fine-tuned by ROS (D’Autréaux and Toledano, 2007; Holmström and Finkel, 2014; Hopkins, 2016a and 2016b). In fact, the clinical application of ROS as signalling molecules is still far away from being an effective solution as a translational therapy (Kamat et al., 2008; Carvalho et al., 2017). Thus, the success of antioxidant therapies depends on considering aspects of redox biology such as its pleiotropism or its cellular and subcellular origin (Juránek et al., 2013; Zhang et al., 2016). In vitro studies using brain cells, principally neurons and astrocytes, have been essential for understanding the highly distinctive biology of these cells. In this sense, it has been shown how these brain cells present different molecular specialization, highlighting the importance of crosstalk during neuron-astrocyte coupling which ensures brain bioenergetic and redox homeostasis (Fernandez‑Fernandez et al., 2012; Bolaños, 2016). Regarding the study of ROS in brain cells, there is recent evidence indicating that the levels of mitochondrial reactive oxygen species (mROS) are immensely greater–about one order of magnitude–in astrocytes than in neurons (Lopez-Fabuel et al., 2016). This finding reiterates the relevance of considering the cellular origin of neural ROS as an important new factor in the study of redox biology in the brain.

To address this issue, we have reported the use of a new method to quantify the levels of mROS in adult mice astrocytes ex vivo, independently of neurons (Vicente-Gutierrez et al., 2019). This methodology has allowed us to demonstrate an effective and specific down-modulation of astrocytic mROS in a transgenic mouse model expressing a mitochondrial-tagged form of catalase (mitoCAT or mCAT) (Vicente-Gutierrez et al., 2019). In this particular study, we showed that astrocytic mROS has an impact on neuronal function and survival by regulating bioenergetics and redox metabolism. Thus, by decreasing endogenous levels of astrocytic mROS, we found that astrocytic mROS may modulate glucose utilization and neuronal function in behaving mice (Vicente-Gutierrez et al., 2019). Using the methodology that we herein describe, we postulated that endogenous mROS in astrocytes play a physiological role in the maintenance of brain homeostasis. Furthermore, in this same study (Vicente-Gutierrez et al., 2019), we were able to characterize a novel conditional mCAT mouse that could be useful for testing the implications of mROS in a desired specific tissue type or pathological model.

Since the study of ROS depends not only on concentration but also on its spatiotemporal distribution, real-time imaging of ROS, possibly in vivo, has become necessary in order for scientists to determine which of their biological activities may present a potential for clinical translation. However, this objective is still unachievable owing to current available techniques (Maulucci et al., 2016). Here, we describe in further detail the protocol previously used in Vicente-Gutierrez et al. (2019) to measure mROS abundance, specifically superoxide anion-abundance in astrocytes ex vivo. Although there are different probes commercially available, the most widely used is MitoSOX^TM. The MitoSOX Red mitochondrial superoxide indicator enters into mitochondria, where it accumulates in response to the mitochondrial membrane potential (ΔΨ_m) and becomes oxidized. This probe has been widely used in astrocytes in culture (Sheng et al., 2013; Angelova et al., 2015) and in cultured hippocampal sections (Ishii et al., 2017). MitoSOX fluorescence intensity is commonly assessed by microscope imaging or flow cytometry. As mentioned above, this fluorescent dye requires an active mitochondrial membrane potential to enter into mitochondria. This implies that to measure mROS, it is necessary to use live cells and to determine, in parallel, the ΔΨ_m in order to confirm that differences in the MitoSOX signal are independent of ΔΨ_m. However, this fact presents a challenge when working with fixed-cells. For instance, this exclude the use of imaging techniques like immunohistofluorescence in fixed-cells which, in addition, to study individual cellular phenomena required the use of different cell markers. To overcome the cell origin problem, others have used genetically encoded probes to assess redox histology in mouse (Fujikawa et al., 2016). Thus, immunohistochemical strategies still offer a suitable approach for detecting the footprints of redox stress. Nowadays, two-photon laser scanning microscopy has become an interesting tool for studying cellular parameters in awake mice, in vivo. However, this technology is not easily accessible in every laboratory and is limited to studying superficial brain areas (Pérez-Alvarez et al., 2013; Wang et al., 2017). Another drawback to this technique is the large amount of time required for sample preparation and for obtaining replicas. In contrast, the protocol that we will describe is relatively easy, accessible, fast and robust.

The ability of certain adeno-associated virus (AAV) vectors to cross the blood-brain barrier after intravenous injection makes it possible to obtain transgene expression in brain cells. The AAV-PHP.eB capsid is one of the serotypes that is able to efficiently transduce the central nervous system (Chan et al., 2017) and is useful for many applications. For instance, we have used it to express the green fluorescent protein (GFP) under the control of the astrocyte-specific glial fibrillary acidic protein (GFAP) short promoter (gfa-ABC₁D) (Lee et al., 2008). Astrocytes were specifically tagged in vivo by infecting mice intravenously through the retro-orbital venous sinus with the AAV-PHP.eB-gfa-ABC₁D-GFP construct. Then, we obtained a single brain cell suspension and adapted a well-known protocol to measure MitoSOX by flow cytometry. Therefore, this protocol combines the use of accessible technology like flow cytometry with the use of commercial probes. Accordingly, another advantage to this approach is its versatility, since it is also useful for assessing other cell-specific processes in brain function. By using other brain cell-specific promoters to target neurons, microglia or oligodendrocytes, for instance, the same phenomenon could be measured independently or simultaneously in different cell types. Moreover, this protocol allows the same sample to be analysed using different commercially available probes to measure different biological processes. Finally, this tool is greatly useful to study other cellular and subcellular phenomena and can be used at different time points during model lifespan as well as different tissue types. Overall, we believe that this straight-forward protocol for measuring mROS in adult astrocytes is beneficial for studying redox biology in vivo at the cellular level in the brain.