Introduction

Mitochondria are key organelles for the maintenance of cellular homeostasis by producing biomolecules, generating bioenergy, calcium buffering and controlling cell death [1]. Evidence points to mitochondria as the organelle signalling cell danger. Likewise, mitochondria can become damaged in response to stress, ageing or disease, resulting in oxidative stress and ultimately cell death [2]. Depending on the type and extension of stress stimuli, mitochondria can respond differently, presenting several mechanisms that are in charge of mitochondrial quality control (MQC). MQC is a set of pathways involved in managing mitochondrial function coupled to cell homeostasis and comprises different processes: fusion, fission, mitophagy and mitochondrial biogenesis [2,3,4]. In response to mild starvation or disfavouring conditions, fusion occurs to improve ATP production and to avoid mitophagy [4]. Nevertheless, mitochondria must be eliminated whenever malfunctioning mitochondria represent a dangerous source of reactive oxygen species (ROS). Thus, fission is activated along with mitochondrial selective autophagy: mitophagy [5]. Finally, mitochondrial biogenesis generates new mitochondria for maintaining the cellular mitochondrial population pool.

There are two main mitophagy pathways: the ubiquitin-dependent Parkin/PINK1 pathway [3, 6] and the BNIP3/NIX receptor-mediated mitophagy [7]. PINK1/Parkin is the most studied mitophagy pathway and mutations on these genes are associated with familial cases of Parkinson disease. Under physiological conditions, PINK1 is degraded in outer membrane of the mitochondria (OMM). Nevertheless, in response to stress, in particular when mitochondrial membrane potential is reduced, PINK1 accumulates on the OMM, recruiting Parkin from the cytosol into OMM for further phosphorylation [6, 8]. Parkin is a E3 ubiquitin ligase that promotes ubiquitination of OMM proteins, initiating the mitophagy process by recruiting the autophagy-associated adaptors [9]. Thus, mitochondrial accumulation of Parkin leads to their degradation via mitophagy. The uncoupler carbonyl cyanide m-chlorophenylhydrazone (CCCP) the most studied inducer of PINK1/Parkin mitophagy pathway [10].

In response to damaged mitochondrial clearance by mitophagy, mitochondrial biogenesis must occur to maintain mitochondrial population homeostasis. Mitochondrial biogenesis is a tightly controlled process that must coordinate the activity of nuclear and mitochondrial transcription factors for a synchronized mitochondrial protein expression. In the nucleus, peroxisome proliferator-activated receptor γ coactivator 1 (PGC-1α) and nuclear respiratory factor 1 (NRF1) are activated and upregulates the expression of mitochondrial transcription factor A (TFAM), allowing replication of the mitochondrial DNA [11]. Moreover, mitochondrial biogenesis promotes cytoprotection and regulates cell metabolism in response to stress [12, 13]. Although the balance between mitophagy and mitochondrial biogenesis has been widely discussed in review papers as key processes of MQC, few experimental data show the direct association of these two processes [14, 15].

Herein, the effect of carbon monoxide (CO) in MQC of astrocytes is explored. CO is a diatomic molecule, commonly known as a lethal gas and toxic air pollutant. Nevertheless, CO is also an endogenous gasotransmitter produced by haem-oxygenase (HO), a stress response enzyme [16]. CO is an ubiquitous signalling molecule to counteract oxidative stress and injury [16, 17], being associated with homeostasis and cytoprotection in several tissues, including the brain [18]. Because of its potential therapeutic use, CO-releasing molecules (CORMs) have been developed to release CO under a control manner in biological settings. Carbon monoxide releasing molecule-A1 (CORM-A1) is a boron-based CORM, which releases CO in response to physiological conditions (temperature and pH) and has been widely used in experimental models [19,20,21,22,23].

In the central nervous system, CO has been firstly described as a neurotransmitter [24]. Interestingly, the brain contains the highest levels of HO expression, suggesting a key role for CO [25]. In fact, we and others have demonstrated that CO prevents apoptosis in neurons [26,27,28] and astrocytes [13, 20, 29]; while in microglia, CO reduces neuroinflammation [30,31,32,33]. In different in vivo models CO is protective by inhibiting blood brain barrier (BBB) permeabilisation in epileptic seizures [19], preventing cell death in ischemic stroke model [34] and in neonatal hypoxic-ischemic encephalopathy [35] and limiting neuroinflammation in haemorrhagic stroke [36]. Furthermore, CO also regulates autophagy. In fact, CO promotes protection by triggering autophagy in sepsis [37], prevents apoptosis in an autophagic dependent manner in islet cells [38] and in astrocytes [39]. Likewise, CO increases ROS generation and signalling that in turn stimulates autophagy [40].

In summary, CO is cytoprotective by (i) improving mitochondrial metabolism, (ii) stimulating mitochondrial biogenesis and (iii) promoting autophagy. Thus, it is hypothesised that CO maintains mitochondrial homeostasis in order to promote cytoprotection. Herein, we explore whether CO, delivered by CORM-A1, regulates mitochondrial turnover by modulating the balance between mitophagy and mitochondrial biogenesis, which in turn protect astrocytes against oxidative stress-induced cell death.

Material and Methods

If not stated, reagents were supplied from Sigma.

mitoQC Model for Mitophagy Detection

MitoQC mice is a model that allows the determination of mitophagy [41] and mice were kindly provided by Dr Ian Ganley (School of Life Sciences, The University of Dundee, Dundee, Scotland). The mice express the MitoQC reporter that has a mitochondrial localisation signal FIS1 coupled to green fluorescent protein (GFP) and red fluorescent protein (mCherry; RFP). When mitochondrial are delivered for lysosomal degradation via autophagy, the green signal of the GFP is quenched inside de acidic lumen of the lysosome and only the red signal is visible. These red dots are indicative of mitophagy (mitolysosomes).

Primary Culture of Astrocytes

Primary cultures of astrocytes were prepared from 1-day-old Balb/c mouse cortex, or from 1-day-old mitoQC mouse model cortex as described [42]. For each culture, 3 to 4 mice pups (both sexs randomly used) were rapidly decapitated, brain cortex was removed and the meninges were carefully stripped off, then the cortex was washed in ice-cold phosphate-buffered saline (PBS), and mechanically disrupted. Single-cell suspensions were plated in T-flasks (four hemispheres/75 cm2) in high glucose Dulbecco’s minimum essential medium (11965092, Thermo Fisher Scientific) supplemented with 20% (v/v) heat-inactivated foetal bovine serum FBS (26140079, Thermo Fisher Scientific) and 1% (v/v) of 100 U/mL penicillin/streptomycin solution (15140122, Thermo Fisher Scientific). Astrocytes were maintained in a humidified atmosphere of 5% CO2 at 37 °C. After 7 days, the dark phase cells growing on the astrocytic cell layer were separated by vigorous shaking and discarded. Culture medium was replaced twice a week for 3 weeks with gradual reduction in FBS content (2nd week 15% (v/v); 3rd week 10% (v/v)). At the 3rd week the confluent astrocytes were detached by mild trypsinisation using trypsin/EDTA (25200056, Thermo Fisher Scientific) (0.25%, w/v) and sub-cultured with FBS at 10%, in 24- and 6-well plates at 50 × 103 and 100 × 103 cells/well, respectively, until reaching full confluence. Purity of astrocytic primary culture was verified by immunocytochemistry using anti-GFAP antibody (Z0334, DAKO) and fluorescent microscopy (Supplementary Fig. 1). Culture presents about 95% of astrocytic cells positive for GFAP. Primary cultures of astrocytes were used up to 5 passages.

Balb/c mice were purchased from NOVA Medical School (Portugal) or from CIB (Madrid, Spain). For primary culture preparation, mice were rapidly decapitated with minimising suffering procedures. The National Institutional Animal Care and Use Committee (Direção Geral de Alimentação e Veterinária) ethical approval was not required since (1-day-old) new-born pups are sacrificed for the single purpose of tissue collection. Nevertheless, the procedure was approved by NOVA Medical School and CSIC Ethical committees accordingly with relevant international guidelines.

Carbon Monoxide Releasing Molecule A1 (CORM-A1) and CO Solutions

The CO releasing molecule, CORM-A1 [21] was used to deliver CO to cell culture. CORM-A1 (SML-0315, Sigma) stock solutions were prepared in water to a final concentration of 5 mM. Then, the solution was filtrated with 0.2 µM filter and stored at − 20 °C to avoid loss of released CO. For each use, an aliquot was thawed and immediately used. The concentration of CORM-A1 and time points for treatments were selected based on the cytoprotective effects of CO in astrocytes subjected to oxidative stress [39]. CO-depleted inactive form (iCORM-A1) was generated to be used as negative control by initially dissolving CORM-A1 in 0.1 M HCl and then bubbling pure N2 (17502001, Thermo Fisher Scientific) through the solution for 10 min in order to remove the residual CO gas. The solution of iCORM-A1 was finally adjusted to pH 7.4 [21]. Then, the solution was filtrated using a 0.2 µM filter and stored at − 20 °C. For each experiment, an aliquot was thawed and immediately used. For the preparation of CO gas solutions, PBS was saturated by bubbling 100% of CO gas (Linde, Germany) for 30 min to produce 10 mM stock solution. The concentration of CO in solution was determined spectrophotometrically by measuring the conversion of deoxymyoglobin to carbon monoxymyoglobin.

Determination of Mitochondrial Population by Flow Cytometry

Astrocytes were plated in a concentration of 50 × 104 cells/well into the 24-well plate. Cells were treated for 1 and 24 h with and without the mitophagy inhibitor Cyclosporin A (CsA) (30024, Sigma) at 5 μM, the mitophagy inducer (uncoupling agent) carbonyl cyanide chlorophenylhydrazone (CCCP) (C2759, Sigma) at 25 μM and chloramphenicol (Cm) (C0378, Sigma) at 10 µM and CORM-A at 12.5 µM. After astrocyte treatment, cells were trypsinized for 5 min at 37 °C and resuspended in complete medium with 10 nM Mitotracker Deep Red to label mitochondria (MTDR, M22426, Invitrogen) incubated for 15 min at 37 °C followed by cytofluorometric analysis with FACS scan (BD FACSCalibour). MTDR labels mitochondria when they have membrane potential, after labelling, mitochondria can become depolarized that MTDR still remains inside, being an accurate manner to quantify mitochondrial mass. Fluorescence intensity for each incubation time was analysed in 10,000 cells and fluorescence value per cell was determined and plotted against cell number.

Fluorescent Microscopy

Astrocytes were isolated from mitoQC mice. After treatments, cells were fixed for 30 min with 4% paraformaldehyde in PBS and washed three times. Coverslips were mounted onto glass slides with Prolong Diamond mounting medium (P36961, Thermo Fisher Scientific) (with DAPI 1:1000 (4',6-diamidino-2-phenylindole, Thermo Fisher Scientific). Confocal z-stacks were obtained with Leica TCS SP8 STED 3× microscopes (Leica Microsystems). Images were taken every 0.5 mm. The red dots indicative of mitophagy were quantified using an ImageJ 1.44 plug in developed in our lab based on a of co-localisation measurement (GFP-mCherry) and a measurement of the population of mCherry (Mitophagy), selecting the populations using the ScatterPlot tool with the different active fluorophores.

Quantitative-Polymerase Chain Reaction (Q-PCR)

For quantification of mitochondrial DNA, astrocytes were plated in a concentration of 100 × 104 cells/well into the 12-well plates for Q-PCR analysis. Cells were treated for 1 and 24 h with CORM-A at 12.5 µM. Genomic DNA was extracted from astrocytes using “High Pure PCR Template Preparation Kit” (11796828001, Roche Diagnostics, Mannheim, Germany). Q-PCR was used for quantification of mitochondrial DNA (mtDNA), which is an indirect measurement of mitochondrial population. Total cell DNA (containing mitochondrial DNA) was extracted from astrocytes using DNeasy Blood & Tissue Kit kit (69504, QIAGEN). Polymerase chain reaction (PCR) was performed using specific forward and reverse primers (Stab Vida) designed for the mitochondrial cytochrome b gene (5′-TTCATGTCGGACGAGGCTT-3′, 3′-TCCTCATGGAAGGACGTAGC-5′) and for the nuclear GAPDH gene (5′-CCTTCATTGACCTCAACTACAT-3′, 3′-CCAAAGTTGTCATGGATGACC-5′) to be used for controlling the cellular amount in each sample. “Fast Strand DNA Master Plus SYBR Green I” (03003230001, Roche) was used in the experimental run protocol, denaturation programme was 95 °C for 10 min, followed by 45 cycles of 95 °C for 10″, 60 °C for 10″ and 72 °C for 10′.

The results were analysed by relative quantification, with aid of the software light cycle 96 (reference 00 000000 0010225, software version 1.01.00.0045, ROCHE). Mitochondrial population was quantified by the amount of mitochondrial DNA relatively to nuclear DNA (internal control). All results were normalized relatively to the control cell sample without any treatment.

siRNA Transfection

PINK1 expression was knocked down by PINK1 coding siRNA transfection according to manufacturer’s instructions (5480187, Invitrogen, UK). Astrocytes at 40% of confluence were transfected using Lipofectamine™ RNAiMAX and Opti-MEM® medium (13778150, Invitrogen, UK); for 1.9 or 3.9 cm2 of astrocytic culture area, 6 or 12 pmol of siRNA (PINK1 or scramble sequence) was used, respectively. At room temperature, siRNA and culture medium were gently mixed with lipofectamine for forming liposomes, and then astrocytes were transfected in the absence of antibiotics. 24 h, 48 h, and 72 h after transfection PINK1 expression was assessed by western blot assay, knocking down was more efficient at 24 h (Supplementary Fig. 2).

Reverse Transcriptase Quantitative PCR (RT-Q-PCR)

Astrocytes were plated in a concentration of 100 × 104 cells/well into the 12-well plates for RT-Q-PCR analysis. Cells were treated for 1 and 24 h with CORM-A at 12.5 µM. For evaluation of PGC-1α gene expression (Mm01208835_m1, Life Technologies), mRNA was extracted from astrocytes. RNA isolation was performed using a mixture of 500 µl of TRIzol (15596026, Thermo Fisher Scientific) and 100 µl of chloroform (1.02445.1000, Merck). Sample was centrifuged (15 min, 12,000 g, 4 °C) and upper aqueous phase was transferred to a new tube for further precipitation of RNA using 250 µl isopropanol (1096341000, Merck). After centrifugation (15 min, 12,000 g, 4 °C), pellet was washed twice with ethanol (1.00971.1000, Merck) (centrifugation 5 min, 7500 g, 4 °C) and resuspended in pure MilliQ water for further analysis. Then, cDNA synthesis was performed using the Transcriptor High Fidelity cDNA synthesis kit (5081955001, Merck). PCR was performed using specific forward and reverse primers (Stab Vida) designed for PGC-1α, (CACAGATTCAGGCCAGTGCT) (TTCTGGTGCTGCAAGGAGAG). “Fast Strand DNA Master Plus SYBR Green I” (03003230001, Roche) was used in the experimental run protocol; denaturation programme was 95 °C for 10 min, followed by 45 cycles of 95 °C for 10 s, 60 °C for 10 s, and 72 °C for 10 s.

Flow Cytometry for Cell Viability Assessment

Astrocytes were treated with CORM-A1 at a final concentration of 12.5 µM for 1 h, followed by cell death induction with t-BHP (B2633, Sigma) at concentrations from 0 to 320 µM for 18 h. To detect cell death induced by t-BHP, cells were collected by trypsinisation and stained with dye propidium iodide (PI, P3566, Invitrogen) at 1 g/mL. Cells were gated by forward and side scatter. A flow cytometer (BD FACSCanto™ II) was used to analyse cell death-associated parameters. The acquisition was done with Diva software and data analysis was performed with FlowJo® software (10.5.3).

Isolated Mitochondria from Primary Culture of Astrocytes

Astrocyte primary cultures were grown in 175 cm2 t-flask up to confluence, then cells were washed with ice-cold PBS and collected by trypsinisation. Samples were centrifugated at 200 g for 10 min, and cells (pellet) were washed in PBS by centrifugation at 200 g for 10 min. The supernatant was discarded and the pellet (cells) incubated in 3.5 mL of hypotonic buffer (0.15 mM MgCl2, 10 mM KCl, 10 mM Tris–HCL, pH 7.6) at 4 °C for 15 min, After the addition of an equal volume of homogenisation buffer (0.15 mM MgCl2, 10 mM KCl, 10 mM Tris–HCL, 0.4 mM PMSF, 250 mM saccharose, pH7.6) twice concentrated, samples were manually homogenized with Douncer potter. Cell extracts were centrifugated at 900 g for 10 min, followed by supernatant centrifugation at 10,000 g for 10 min. The mitochondrial pellet was resuspended in 110 µL of the homogenisation buffer and the total amount of protein was quantified using BCA assay (Pierce, Illinois). All the steps were carried out at 4 °C.

Protein Extraction and Immunoblotting

Isolated mitochondria were separated under reducing electrophoresis on a 1 mm thick 15% SDS–polyacrylamide (SDS-PAGE) gel. Samples were transferred onto a PVDF membrane (MERK). Membranes were incubated overnight at 4 °C with antibodies against PINK1 protein (GTX107851, GeneTex, dilution 1/1000) and Parkin protein (ab77924, Abcam, dilution 1/1000) and cytochrome c oxidase (COX11, sc-98918, Santa Cruz, dilution 1/1000), used as a loading control. Blots were developed using the ECL (enhanced chemiluminescence, 32106, Thermo Fisher Scientific) detection system after incubation with horseradish peroxidase-conjugated secondary antibodies (GE Healthcare, dilution 1/5000), for 1 h at room temperature.

Statistics

Data are presented as mean ± SEM from at least 3 independent experiments. Reported results were statistically evaluated using GraphPad Prism (GraphPad Prism 9). The normality of the data was tested using D’Agostino and Pearson omnibus normality test. Data that passed the normality test was analysed using unpaired t-test, one-way ANOVA, or two-way ANOVA. The remaining data was analysed using Kruskal–Wallis test or Mann–Whitney test in the appropriated cases. The statistical test and the number of replicas are stated in the figure legend. P < 0.05 was considered significant.

Results

CORM-A1 Promotes Mitophagy

In order to assess whether CO stimulates mitophagy, primary culture of cortex astrocytes isolated from mitoQC mice were treated with CORM-A1 at 12.5 µM for 1 h, 3 h, 6 h, 8 h or 24 h and mitophagy was assessed by counting red dots as the number of mitochondria engulfed by lysosomes per cell. The 12.5 µM concentration of CORM-A1 is the optimal concentration for promoting cytoprotection in primary culture of astrocytes [20, 39]. 1 h of CORM-A1 treatment reaches the highest levels of mitophagy (Fig. 1 A and B). Likewise, mitochondrial population was measured by flow cytometry using Mito tracker deep red (MTDR) dye [43]. In fact, CORM-A1 reduces mitochondrial population after 1 h of treatment, indicating the activation of mitophagy (Fig. 1C). This effect is reverted by co-treatment with cyclosporin A (CsA), which is an indirect inhibitor of mitophagy preventing mitochondrial membrane depolarisation [43] (Fig. 1C). Moreover, the classical inducer of mitophagy CCCP was used as a positive control and reduced mitochondrial population assessed by MTDR. CCCP promotes mitophagy by inducing mitochondrial depolarisation. In contrast, CsA treatment enhanced mitochondrial population, which may indicate inhibition of basal mitophagy (Fig. 1C). In order to control that CORM-A1 effect is dependent on CO release, astrocytes were treated with inactivated CORM-A1 or with CO-saturated PBS solutions, and mitochondrial mass was assessed by MTDR. In fact, inactivated CORM-A1 did not promote mitophagy, while CO solutions promoted mitophagy following 1 h of treatment (Fig. 1D). This data indicates that CO is the factor stimulating mitophagy. In conclusion, CORM-A1 promotes mitophagy at short time period (1 h), which was measured by mitochondria co-localisation into lysosomes (mitolysosomes) and by the reduction of mitochondrial mass.

Fig. 1
figure 1

CORM-A1 induces mitophagy in astrocytes. A Primary culture of astrocytes isolated from mitoQC transgenic pups was pre-treated with 12.5 µM of CORM-A1 for 1 h, 3 h, 6 h, 8 h, and 24 h. Representative picture of astrocyte mitophagy followed by fluorescent microscopy. Mitochondrial protein FIS1 is coupled to GFP and RFP. When mitochondrial are engulfed by lysosomes for degradation, the green signal of the GFP is quenched inside de acidic lumen of the lysosome and only the red signal is visible. Thus, red dots are indicative of mitolysosomes (mitophagy). B Quantification of mitolysosomes per cell; graphs represent the mean ± SEM of 3 biological experiments; data were analysed with unpaired t-test; *p < 0.05 and **p < 0.01. C Primary cultures of astrocytes cultured in 24-well plate were treated with 25 µM CCCP or 12.5 µM of CORM-A1 for 1 h with and without the inhibitor CsA at 5 µM, which was added simultaneously with mitophagy inducers. MTDR was used to determine mitochondrial mass by flow cytometry. The intensity in the FL4 channel was normalized to untreated control cells. Graphs represent the mean ± SEM of five experiments performed in triplicate, analysed with one way ANOVA test. *p < 0.05 and **p < 0.01 compared with control and ##p < 0.01 compared to CORM-A1 treatment. D In order to control the role of CO in the modulation of mitochondrial population and mitophagy, MTDR-based mitochondrial population was performed in the presence of 12.5 µM of inactivated CORM-A1 or 50 µM of CO gas (PBS-saturated solution) for 1 h and 24 h. Graphs represent the mean ± SEM of five experiments performed in triplicate, analysed with the two-way ANOVA test. *p < 0.05 compared to control

Full size image

CORM-A1-Induced Mitophagy Is Dependent on PINK1/PARKIN Pathway

During mitophagy stimulation, in particular mitochondrial inner membrane depolarisation or mitochondrial stress, PINK1 accumulates in the outer mitochondrial membrane (OMM), recruiting and activating cytosolic Parkin. Astrocytes were treated with CORM-A1 and CCCP (positive control) for 1 h, and enriched mitochondrial fraction was isolated from astrocytic cell culture for further PINK1 and Parkin detection by Western blot (Fig. 2 A, B and C). In fact, CORM-A1 increased PINK1 and Parkin accumulation in mitochondrial fraction (Fig. 2 B and C). These data might indicate PINK1/Parkin pathway involvement in CORM-A1-induced mitophagy. Likewise, whenever PINK1 expression was knocked down (KD PINK1) in astrocytes isolated from the mitoQC mice, there was a decrease on basal and CORM-A1-induced mitophagy, which was assessed by the number of mitolysosomes (co-localisation of mitochondria and lysosomes) per cell (Fig. 2 D and E). Thus, basal and CORM-A1-induced mitophagy are dependent on PINK1 expression.

Fig. 2
figure 2

CORM-A1-induces PINK1/Parkin-dependent mitophagy in astrocytes. A Mitochondrial fractions isolated from primary culture of astrocytes treated with CORM-A1 at a final concentration of 12.5 µM or with CCCP at 25 µM (positive control) during 1 h were analysed by immunoblotting using antibodies against PINK1 and Parkin and normalized using anti-COX11. B, C Quantification of the presence of PINK1 (B) and Parkin (C) in mitochondrial enriched fraction. Graphs represent the mean ± SEM of four experiments analysed with Mann–Whitney test, with *p < 0.05. D Primary culture of astrocytes isolated from mitoQC transgenic animals and with PINK1 expression knocked down with siRNA (10 pmol) was pre-treated with 12.5 µM of CORM-A1 for 1 h. Representative picture of astrocyte mitophagy assessed fluorescent microscopy. E Quantification of mitolysosomes (red dots) per cell (number of cells per condition ≥ 20). Graphs represents mean ± SEM. Statistic was made using two-way ANOVA test with *p < 0.05 and **p < 0.01 compared to control without treatment, and ##p < 0.01 compared to CORM-A1 treatment

Full size image

CORM-A1 Promotes Mitochondrial Biogenesis and Modulates Mitochondrial Turnover

After 24 h of CORM-A1 treatment, we did not find stimulation of mitophagy measured by mitoQC reporter in astrocytes (Fig. 3A and B). In fact, mitophagy levels decreased over time following CORM-A1 stimulation (Fig. 1 A and B). Mitophagy can also be indirectly assessed by mitochondrial mass using MTDR dye. While CORM-A1 decreased mitochondrial mass after 1 h, which indicates mitophagy stimulation (Fig. 1C); following 24 h mitochondrial mass is back to control levels (Fig. 3C). This data suggests that as late response CORM-A1 may stimulate mitochondrial biogenesis for the reestablishment of normal mitochondrial population. In contrast, the positive control CCCP still decreased mitochondrial population at 24 h after treatment, indicating that at 24 h CCCP maintains mitophagy stimulation (Fig. 3C). Chloramphenicol (Cm) that prevents protein synthesis and inhibits mitochondrial biogenesis was used to assess whether CORM-A1-induced variations on mitochondrial mass are dependent on mitophagy or mitochondrial biogenesis. In fact, astrocytes co-treated with CORM-A1 and Cm for 24 h presented lower levels of mitochondrial mass, similar to CCCP treatment (Fig. 3C). Thus, it suggests that CORM-A1 may stimulate mitochondrial biogenesis that is abolished by Cm treatment to block protein synthesis. In contrast, astrocytes treated only with Cm presented mitochondrial population levels similar to control. It can be speculated that basal levels of mitochondrial biogenesis are lower, since they are not affected by 24 h of chloramphenicol.

Fig. 3
figure 3

CORM-A1 induces mitochondrial biogenesis. A Primary culture of astrocytes isolated from mitoQC transgenic mice pups was pre-treated with 12.5 µM of CORM-A1 for 24 h. Representative picture of astrocytic mitophagy followed by fluorescent microscopy. B Quantification of mitolysosomes (red dots) per cell. Graph represents the mean ± SEM of 3 experiments performed with 4 technical replicas, and analysed with unpaired t-test with *p < 0.05. C Primary cultures of astrocytes cultured in 24-well plate were treated with 25 µM CCCP or 12.5 µM of CORM-A1 for 24 h, with and without the mitophagy inhibitor CsA at 5 µM or the protein synthesis inhibitor chloramphenicol (Cm) at 10 µM; both inhibitors were added simultaneously with mitophagy inducers. MTDR was used to determine mitochondrial mass by flow cytometry. The intensity in the FL4 channel was normalized to untreated control cells. Graphs represent the mean ± SEM of five experiments performed in triplicate, analysed with the one-way ANOVA test. D Primary cultures of astrocytes cultured in 24-well plate were treated with 12.5 µM of CORM-A1 and 25 µM of CCCP for 1 h and 24 h, followed by DNA extraction for measuring mitochondrial cytochrome b (mtCyt b) gene to assess mitochondrial DNA amount, which is represented by fold increase when compared to control. Graphs represent the mean ± SEM of four experiments performed in triplicate; data were analysed with two-way ANOVA test. E Primary cultures of astrocytes cultured in 24-well plate were treated with 12.5 µM of CORM-A1 for 1 h and 24 h, followed by mRNA extraction for measuring mitochondrial PGC-1α gene to assess mitochondrial biogenesis, which is represented by fold increase when compared with control. All values are mean ± SEM of five experiments performed in triplicate and data were analysed with Bonferroni’s multiple comparison test. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001 compared with control

Full size image

In order to further verify the role of CORM-A1 in mitochondrial biogenesis, mitochondrial DNA was measured using cytochrome b DNA quantification as mitochondrial marker. CORM-A1 increased mitochondrial DNA levels in both time points: 1 h and 24 h (Fig. 3D). Thus, it demonstrates that CO induces mitochondrial DNA replication early at 1 h, suggesting that mitochondrial biogenesis stimulation may occur while mitophagy is still activated. Interestingly, the classical mitophagy inducer CCCP also promoted mitochondrial DNA replication at 1 h and 24 h, but presented a much lower increase (Fig. 3D). In conclusion, CO-induced mitochondrial biogenesis appears to occur simultaneously with mitophagy activation.

To assess the direct CO effect on mitochondrial biogenesis in a more accurate manner, we analysed the expression of PGC-1α, as this factor upregulates the expression of mitochondrial transcription factor A (TFAM), and is directly implicated in mitochondrial biogenesis. Accordingly with mitochondrial DNA replication, CORM-A1 increases PGC-1α expression following 1 h (Fig. 3E). Nevertheless, PGC-1α expression goes back to basal levels after 24 h (Fig. 3E). One may speculate that at 24 h stimulation of mitochondrial biogenesis is already reverted to basal levels.

In summary, all these data indicate that at 1 h CORM-A1 stimulates mitophagy (Fig. 1) and simultaneously promotes the activation of compensatory mitochondrial biogenesis, assessed by increased levels of mitochondrial DNA (Fig. 3D) and PGC-1α upregulation at 1 h (Fig. 3E) in order to reach mitochondrial population homeostasis at 24 h. Finally, one can still speculate that the low levels of mitolysosomes found at 24 h after CORM-A1 treatment (Fig. 1 A and B) indicate that activated mitochondrial biogenesis leads to new and functional mitochondria, which in turn decreases the need of mitophagy clearance of dysfunctional mitochondria.

Taken all together, these data suggest that CO is an important player in the turnover of mitochondria by promoting mitophagy and mitochondrial biogenesis, two key processes of mitochondrial quality control. In fact, mitochondrial quality control is not limited to the elimination of dysfunctional mitochondria by mitophagy, but it must also promote generation of new mitochondria. Thus, those two processes might be interconnected and interdependent.

CORM-A1-Induced Mitochondrial Biogenesis Is Dependent on Mitophagy

In order to assess whether the mitophagy and mitochondrial biogenesis are interdependent, mitophagy was genetically inhibited and mitochondrial biogenesis was measured. PINK1 expression was knocked down by siRNA treatment for 24 h and mitochondrial biogenesis was assessed by quantification of PGC-1α expression following 1 h of CORM-A1 treatment. In fact, knocking down PINK1 reverted CORM-A1-induced expression of PGC-1α (Fig. 4A). Therefore, CORM-A1-induced mitochondrial biogenesis is dependent on PINK1-dependent mitophagy.

Fig. 4
figure 4

CORM-A1-induced mitochondrial biogenesis and CORM-A1 cytoprotection are dependent on mitophagy. A Primary cultures of astrocytes cultured in 24-well plate were treated with PINK1 siRNA for knocking down its expression. After 24 h, astrocytes were treated with 12.5 µM of CORM-A1 for 1 h, followed by mRNA extraction for measuring mitochondrial PGC-1α gene to assess mitochondrial biogenesis, which is represented by fold increase when compared with control. All values are mean ± SEM (n = 5) and data were analysed with the Mann–Whitney test. **p < 0.01 and #p < 0.05 compared with CORM-A1 treatment. B Cell viability was assessed by flow cytometry, using propidium iodide that dyes permeable plasma membrane, which are considered dead cells. Astrocytic PINK1 expression was knocked down for 24 h; then, astrocytes were treated with 12.5 µM of CORM-A1 for 24 h, followed by challenging them to die with the pro-oxidant t-BHP (0 to 320 µm) for 18 h. All the values are mean ± SEM (n = 3) and data were analysed with the Mann–Whitney test. *p < 0.05 compared to control and #p < 0.05 compared with CORM-A1 treatment

Full size image

CORM-A1 Cytoprotection Depends on Mitochondrial Quality Control

Several evidence demonstrate that autophagy is implicated in CO-induced protection and prevention of cell death [37,38,39]. Likewise, overexpression of HO-1 increases mitophagy and mitochondrial biogenesis [44]. Thus, the potential association of CO modulation of mitochondrial quality control with prevention of cell death was evaluated by genetic inhibition of mitophagy. CORM-A1-treated astrocytes presented higher viability following oxidative stress-triggered cell death with the pro-oxidant tert-butylhydroperoxide (tBHP) (Fig. 4B). This is in accordance with previously data [13, 20, 29, 39]. Mitophagy was prevented by knocking down PINK1 expression with siRNA transfection for 24 h. In PINK1 KD astrocytes, the cytoprotection role of CORM-A1 against oxidative stress was reverted (Fig. 4B). In conclusion, mitophagy is an important cellular process for CORM-A1 to be cytoprotective in primary astrocytes.

Discussion

Astrocytes are key glial cells in the maintenance of brain homeostasis by supporting neuronal function [45] and by modulating metabolism in the neurovascular unit [46]. Thus, targeting astrocytes and promoting astrocytic homeostasis and cytoprotection may be a promising strategy to protect neurons. Previously, it was demonstrated that CO inhibits astrocytic cell death triggered by oxidative stress via different cellular mechanisms, namely via ROS signalling [29], improvement of mitochondrial metabolism [13] or autophagy activation [39]. Here, we show that CO confers cytoprotection against oxidative stress-induced cell death in astrocytes in a mitophagy-dependent manner. Moreover, CO promotes both mitophagy and mitochondrial biogenesis in an inter-dependent manner. Actually, CO-induced mitochondrial biogenesis is dependent on mitophagy, since mitochondrial biogenesis decreased when PINK1 expression was knocked down. One hour of CO treatment promotes mitophagy in a PINK1/Parkin-dependent manner. Then, mitophagy levels decrease over time, being the re-establishment of normal mitochondrial population at 24 h after CO exposure. Nevertheless, CO-stimulated mitochondrial biogenesis appears to start at 1 h of treatment, which was assessed by the increased levels of PGC-1α expression and mitochondrial DNA replication. Therefore, one can speculate that clearance of mitochondria triggers generation of new mitochondria, since it was shown that mitophagy activation occurs simultaneously with mitochondrial biogenesis. Likewise, the low levels of mitolysosomes found at 24 h after CORM-A1 treatment may indicate control mechanisms to stop the process where mitochondrial turnover with generation of new and functional mitochondria inhibits basal mitophagy.

Finally, knocking down PINK1 expression prevents CO-induced mitochondrial biogenesis assessed by PGC-1α expression, indicating that mitochondrial biogenesis is dependent on PINK1-dependent mitophagy. The interdependence between mitophagy and mitochondrial biogenesis has been widely discussed in the literature, namely in review papers. Nevertheless, few experimental data demonstrate the direct link of these two processes. In SH-SY5Y neuroblastoma cell line, CCCP-induced mitophagy promotes Nrf2 and TFEB translocation into the nucleus, which in turn promotes mitochondrial and lysosomal biogenesis [47]. In the same model, Parkin overexpression upregulates TFAM that is involved in mitochondrial DNA replication [48]. In PC12 cell line, under basal conditions, knocking down PINK1 expression decreased expression of PGC-1α and TFAM, indicating that mitophagy control also modulates mitochondrial biogenesis [49]. In contrast, under rotenone-induced stress and mitophagy, PINK1/Parkin pathway and PGC-1α expression present mutual antagonism for controlling the balance between mitophagy and mitochondrial biogenesis and maintaining mitochondrial population [49]. Taken together, these data indicate that mitophagy and mitochondrial biogenesis may mutually compensate and are interdependent, being the both processes part of MQC.

Further work is needed for disclosing other molecular players than PINK1. A potential candidate to regulate both processes is the protein PARIS, which appears to inter-connect Parkin and PGC-1α. In fact, PARIS is a transcriptional repressor (member of the family of KRAB zinc-finger proteins), which is ubiquitinated by Parkin targeting it for proteasomal degradation [50, 51]. Moreover, PARIS is a physiological transcriptional repressor of the PGC-1α. Thus, PARIS degradation upregulates PGC-1α and subsequently the expression of PGC-1-dependent nuclear respiratory factors genes, promoting therefore mitochondrial biogenesis [52].

Low concentrations of exogenous CO or endogenous CO are cytoprotective by stimulating mitochondrial biogenesis in different models [12, 13, 53], as well as in astrocytes in a PGC-1α-dependent manner [54]. Likewise, CO also promotes cytoprotection in an autophagy dependent manner [37,38,39]. Concerning HO-1 activity, its overexpression stimulates mitophagy and mitochondrial biogenesis in cardiomyocytes [44]. Herein, we have demonstrated that CORM-A1 stimulates both mitophagy and mitochondrial biogenesis, being the last one dependent on mitophagy. Recently, it was also shown that CO confers protection to hepatocytes by simultaneously stimulating mitophagy and mitochondrial biogenesis in a TFEB dependent manner [55]. Thus, CO emerges as a key endogenous molecule controlling MQC and improving mitochondrial homeostasis.

Several underlying molecular mechanisms have been described in the literature concerning how endogenous or low concentrations of exogenous CO target mitochondria to promote cytoprotection. CO generates low amounts of mitochondrial ROS, which as signalling molecules confer anti-proliferative [56], anti-apoptotic [26, 29] or anti-inflammatory effects [57]. Regarding mitochondrial metabolism, CO increases oxidative phosphorylation, which protects against cell death [13], improves chemotherapy response [58] and enhances neuronal differentiation [59]. Mitochondrial population and mitochondrial biogenesis is stimulated by CO, which confers protection in cardiomyocytes [53], in astrocytes [13, 54] or in hepatocytes [60]. Therefore, at molecular level, it can be hypothesized that CO-induced ROS generation facilitates mitophagy activation for the clearance of old/dysfunctional mitochondria, which in turn promotes biogenesis and the renewing of mitochondrial population. Under physiological conditions CO may facilitate basal and normal mitochondrial turnover. While, under oxidative stress, CO enhances the elimination of damaged mitochondria by mitophagy.

In summary, CO prevents astrocytic cell death by the maintenance of mitochondrial homeostasis. In fact, CO improves mitochondrial clearance and turnover by simultaneously stimulating mitophagy and mitochondrial biogenesis in an inter-dependent manner. This data opens new windows for potential future clinical cytoprotective strategies based on CO-mediated mitochondrial quality control.