Abstract
Mitochondrial fission catalyzed by dynamin-related protein 1 (Drp1) is necessary for mitochondrial biogenesis and maintenance of healthy mitochondria. However, excessive fission has been associated with multiple neurodegenerative disorders, and we recently reported that mice with smaller mitochondria are sensitized to ischemic stroke injury. Although pharmacological Drp1 inhibition has been put forward as neuroprotective, the specificity and mechanism of the inhibitor used is controversial. Here, we provide genetic evidence that Drp1 inhibition is neuroprotective. Drp1 is activated by dephosphorylation of an inhibitory phosphorylation site, Ser637. We identify Bβ2, a mitochondria-localized protein phosphatase 2A (PP2A) regulatory subunit, as a neuron-specific Drp1 activator in vivo. Bβ2 KO mice of both sexes display elongated mitochondria in neurons and are protected from cerebral ischemic injury. Functionally, deletion of Bβ2 and maintained Drp1 Ser637 phosphorylation improved mitochondrial respiratory capacity, Ca2+ homeostasis, and attenuated superoxide production in response to ischemia and excitotoxicity in vitro and ex vivo. Last, deletion of Bβ2 rescued excessive stroke damage associated with dephosphorylation of Drp1 S637 and mitochondrial fission. These results indicate that the state of mitochondrial connectivity and PP2A/Bβ2-mediated dephosphorylation of Drp1 play a critical role in determining the severity of cerebral ischemic injury. Therefore, Bβ2 may represent a target for prophylactic neuroprotective therapy in populations at high risk of stroke.
SIGNIFICANCE STATEMENT With recent advances in clinical practice including mechanical thrombectomy up to 24 h after the ischemic event, there is resurgent interest in neuroprotective stroke therapies. In this study, we demonstrate reduced stroke damage in the brain of mice lacking the Bβ2 regulatory subunit of protein phosphatase 2A, which we have shown previously acts as a positive regulator of the mitochondrial fission enzyme dynamin-related protein 1 (Drp1). Importantly, we provide evidence that deletion of Bβ2 can rescue excessive ischemic damage in mice lacking the mitochondrial PKA scaffold AKAP1, apparently via opposing effects on Drp1 S637 phosphorylation. These results highlight reversible phosphorylation in bidirectional regulation of Drp1 activity and identify Bβ2 as a potential pharmacological target to protect the brain from stroke injury.
Introduction
Bβ2 is a member of a large family of regulatory subunits of the protein phosphatase 2A (PP2A) heterotrimer complex, which also contains a scaffolding and catalytic subunit. An alternative splice product of the gene PPP2R2B (Fig. 1-1A), a trinucleotide expansion in which causes spinocerebellar ataxia type 12 (Holmes et al., 1999), Bβ2 expression is neuron specific but widespread throughout the nervous system (Dagda et al., 2003). During neuronal stress in vitro, Bβ2 targets PP2A to the outer mitochondrial membrane, promoting mitochondrial fission and cell death through dephosphorylation of the mitochondrial fission enzyme dynamin-related protein 1 (Drp1) at the inhibitory phosphorylation site serine 637 (numbering according to human splice isoform 1; Dagda et al., 2003, 2005, 2008; Merrill et al., 2013).
Phenotypic characterization of Bβ2 KO mice. A, Weight gain on standard chow of mice of the indicated sex and genotype was monitored over 10–12 weeks. B, Motor coordination and motor learning was tested with the accelerating Rotarod test (4–40 rpm over 4 min). C, D, Time spent in the periphery versus center of an open field was assessed as a measure of anxiety. C, Representative paths of individual mice, and (D) percentage time spent in the periphery. Tests revealed sex difference in weight, but neither sex nor genotype differences in other measures. Also unchanged by the Bβ2 KO were neuromuscular strength as measured by the inverted wire hang test. Also, no hindlimb clasping was observed. Plotted are means ± 95% CI of 4–5 mice per sex and genotype in A, and 11–13 per genotype in B and D. Bβ2 KO strategy and validation are shown in Figure 1-1. Additional open-field and home-cage activity data are shown in Figure 1-2.
Figure 1-1
Bβ2 deletion strategy and confirmation of deletion in mice. A, Schematic illustrating alternative splice products of the gene PPP2R2B and location of CAG repeats associated with spinocerebellar ataxia type 12 (SCA12). B, Schematic illustrating specific targeting of exon 1.2, which encodes the divergent N-terminal tail of the Bβ2 splice isoform. C, mRNA expression of Bβ2 in the forebrain in Bβ2+/+ and Bβ2−/− mice, n = 3 mice/genotype. D, Representative Western blot for Bβ2 and other PP2A heterotrimer subunits in forebrain homogenate from mice of the indicated genotypes. Download Figure 1-1, TIF file.
Figure 1-2
Additional behavioral characterization of Bβ2 KO mice. A–C, Additional measures from a 20 min open field reveal no effect of the Bβ2 KO on total distance traveled (A), speed (B), and the number of times a mouse reared itself on its hindpaws (C). D–F, Results from home cage activity monitoring via infrared camera (Noldus PhenoTyper) show that the Bβ2 KO does not influence diurnal activity patterns and total activity over a 23 h monitoring period in terms of total distance traveled (D), mobile periods (E), and time spent near the food supply (F). Shown are mean ± 95% CI in all bar graphs and mean ± SEM in the x–y graphs of D–F of 9–11 mice per genotype in A and B, 10 mice/genotype in C, and 7–9 mice/genotype in D–F. There are no significant changes according to unpaired, two-tailed t tests. Download Figure 1-2, TIF file.
Figure 1-3
Statistical tests and statistical values. DF, degrees of freedom. Download Figure 1-3, DOC file.
As we have reported recently, phosphoregulation of Drp1-S637 plays an important role in determining ischemic sensitivity (Flippo et al., 2018). Deletion of the outer mitochondrial scaffolding protein A-kinase anchoring protein 1 (AKAP1) in mice reduced inhibitory phosphorylation of Drp1-S637 at the mitochondria, promoted Drp1 localization to mitochondria and mitochondrial fragmentation, and exacerbated cerebral ischemic damage following transient middle cerebral artery occlusion (MCAO; Flippo et al., 2018). These results are in line with prior work, showing that pharmacological inhibition of Drp1 with the small molecule mdivi-1 is protective in rodent models of ischemic stroke (Grohm et al., 2012; Zhao et al., 2014; Li et al., 2015). However, the specificity, efficacy, and mechanism by which mdivi-1 protects against ischemic damage has recently been challenged (Bordt et al., 2017). Therefore, whether inhibition of Drp1 is protective against stroke in vivo remains an open question.
To address this question, we have deleted Bβ2 in mice and examined the role of inhibitory Drp1 S637 phosphorylation in ischemic sensitivity in vivo. We found that deletion of even a single Bβ2 allele significantly decreased infarct volume 24 h after transient MCAO. Furthermore, loss of Bβ2 promoted mitochondrial elongation specifically in neurons without altering total mitochondrial mass. Consistently, deletion of Bβ2 decreased localization of the catalytic subunit of PP2A to the mitochondria, leading to hyperphosphorylation of mitochondrial Drp1 at S637 due to unopposed AKAP1/PKA. Mechanistically, we found that deletion of Bβ2 delays calcium dysregulation in the CA1 region of the hippocampus after challenge with oxygen-glucose deprivation (OGD). Furthermore, OGD failed to induce a surge in superoxide production in hippocampal slices from Bβ2 KO mice. To examine cell autonomy, we cultured hippocampal neurons from mice lacking Bβ2 and observed similar delayed Ca2+ excitotoxicity and decreased superoxide production in response to glutamate excitotoxicity. Consistent with resilience to metabolic stress, respirometry revealed increased spare respiratory capacity in Bβ2 KO neurons. Introducing a phosphomimetic Drp1 mutant into hippocampal neurons recapitulated all outcome improvements observed in neurons lacking Bβ2, establishing Drp1 Ser637 as a critical substrate in ischemic sensitization by outer-mitochondrial PP2A. Last, deleting Bβ2 in mice lacking AKAP1 rescued worsened stroke outcomes observed in mice that only lack AKAP1, indicating convergent mechanisms in vivo.
Materials and Methods
Animals
All animal work was performed in accordance with the guidelines of the animal ethics committee of the University of Iowa. Mice were group-housed in a colony maintained with a standard 12 h light/dark cycle and given food and water ad libitum. Experiments were performed on age-matched mice of both sexes except for MCAO and TEM experiments in which only male mice were used. Experiments were conducted according to the Guide for the Care and Use of Laboratory Animals, as adopted by the National Institutes of Health, and with approval of the University of Iowa, Institutional Animal Care and Use Committee.
In our Bβ2-null mice, the N-terminal 24 aa (exon 1) of Bβ2 have been replaced with the coding sequence for βgal (lacZ) and a neomycin cassette surrounded by two loxP sites followed by a polyadenylation signal to terminate the transcript. This selectively deletes the Bβ2 splice variant. Targeted ES cells were injected into C57Bl/6 blastocysts, which were implanted into pseudo-pregnant ICR strain females to create chimeric mice (Gene Targeting Core Facility; Baoli Yang, director). Once germline transmission was established, a Bβ2 +/− mouse was crossed with a transgenic mouse carrying the EIIa-Cre gene (which expresses Cre recombinase after fertilization) to remove the neomycin cassette and the offspring were backcrossed to C57BL/6 mice to generate Bβ2 +/− mice without the EIIa-Cre transgene. To distinguish Bβ2 +/+, +/−, and −/− by PCR, the following PCR primers were used: ACT CCA GTG CAA CAA CCG GCA CTC C (Rev), CCT TTG GAA GAT GAA ATG CTT CTC TCG (F 1), and GGC GCG CCT TAA TTA AGG ATC CTG C (F 2). The expected PCR product size from wild-type (WT) mice is ∼600 bp, whereas the null allele gives rise to a product of ∼400 bp.
Antibodies and reagents
A mouse monoclonal antibody recognizing Bβ2 was generated at the University of Iowa Hybridoma facility. Briefly, a peptide corresponding to part of the unique N-terminus of Bβ2 (aa 3–20, CFSRYLPYIFRPPNTILS) was synthesized and conjugated via the N-terminal Cys to maleimide-activated bovine serum albumin (ThermoFisher). Hybridomas were isolated and screened by ELISA for antibody secretion according to standard protocols. The rabbit polyclonal antibody specific for Bβ1 was described previously (Strack et al., 1998). The rabbit polyclonal antibody directed against MnSOD was a kind gift from Frederick Domann (University of Iowa). The following commercially available antibodies were used: rabbit anti-phospho-Ser637 Drp1 (Cell Signaling Technology), mouse anti-PP2A/C, mouse anti-Drp1, mouse anti-Opa1, mouse anti-MAP2B (BD Transduction Laboratories), mouse anti-beta-tubulin (Developmental Studies Hybridoma Bank, University of Iowa), rabbit anti-TOM20, rabbit anti-TOM40 (Santa Cruz Biotechnology), mouse anti-Hsp60 (Proteintech), mouse anti-OXPHOS, rabbit anti-VDAC (Abcam), infrared fluorophore-coupled secondary antibodies (Licor), and HRP-conjugated secondary antibodies (PerkinElmer). Hoechst 33 342, dihydroethidium (DHE), and Lipofectamine 2000 were from Invitrogen. All other reagents were obtained from Sigma-Aldrich.
Viral vectors
AAV5-CAG-GCaMP6f-WPRE was a kind gift from Dr. Catherine Marcinkiewcz and was used as described in the Stereotactic injection section. Lentivirus expressing phosphomimetic Drp1 (S637D) and outer mitochondria-targeted GFP (Mas70-GFP) were described previously (Flippo et al., 2018). Vectors express the major neuronal Drp1 splice variant, which includes all three alternative coding exons (3, 16, and 17; Uo et al., 2009; Strack et al., 2013). To replace endogenous with mutant Drp1, the lentivirus also expresses a Drp1-targeting shRNA from the H1 promotor inserted into the MfeI site of the vector pFIV3.1-CAGmcs. Virus was generated at the University of Iowa Viral Vector Core. Viral infection of cortical and hippocampal cultures was performed at 7 DIV by removing one-half of the media volume from each well and replacing with NB-A media containing virus (1:100–1:250 dilution for respirometry, 1:1000 for Ca2+/DHE imaging). After 6 h incubation, virus-containing media was replaced with conditioned NB-A media.
Middle cerebral artery occlusion
This procedure was performed at the University of Iowa and at Legacy Research, both of which are accredited by the Association for Assessment and Accreditation of Laboratory Animal Care in accordance with protocols approved by the Institutional Animal Care and Use Committee of each facility, as well as the principles outlined in the National Institute of Health Guide for the Care and Use of Laboratory animals. A murine focal ischemia model was used to determine the potential protective effects of Bβ2 knock-out. In this model, a 45 min period of middle cerebral artery occlusion results in significant infarction of the ipsilateral hemisphere, as determined by staining with the vital dye triphenyltetrazolium chloride (TTC). Adult mice were anesthetized using 5% isoflurane, and then maintained at 2% isoflurane. Depth of anesthesia was determined by periodic (every 10 min) assessment of respiration and pinch withdrawal reflex. Brain temperature was measured using a temporalis muscle-implanted thermocouple (Omega) and controlled within the range of 37 ± 0.5°C. The middle cerebral artery (MCA) was occluded for a 45 min period by threading a silicone-coated 6-0 monofilament nylon surgical suture through the external carotid to the internal carotid, and blocking its bifurcation into the MCA and anterior cerebral artery (Traystman, 2003). Achievement of ischemia and reperfusion was confirmed by monitoring regional cerebral blood flow (CBF) in the area of the right middle cerebral artery, recording CBF 2 min after insertion and 5 min after removal of the suture. CBF was monitored through a disposable microtip fiber-optic probe (diameter 0.5 mm) connected through a Master Probe to a laser Doppler computerized main unit (PF5001, Perimed) and analyzed using PSW Perisoft 2.5 (Kawano et al., 2006) for each mouse, which showed that regional cerebral blood flow was reduced by 70–90% (Fig. 2-1).
Figure 2-1
Cerebral blood flow after MCAO and after reperfusion. Brain perfusion for the experiments shown in Figure 2, A and B, was quantified by laser Doppler flowmetry before and 2 min after occlusion of the middle cerebral artery and 5 min after removal of the filament (reperfusion). Blood flow is expressed as percentage initial (before occlusion) and shown as mean ± 95% CI. No significant differences between genotypes were found (two-way ANOVA). Download Figure 2-1, TIF file.
Bβ2 KO decreases infarct volume following MCAO and increases mitochondrial profile area in neurons. A, Triphenyltetrazolium chloride staining of sliced brains of WT (+/+) and Bβ2 KO (−/−) mice 24 h following 45 min MCAO; red represents living tissue, whereas white defines the infarcted area. B, Infarct volume quantification; shown are data points for individual mice (5–9/genotype) as well as mean ± 95% CI. C, TEM images of hippocampal neurons in brains of WT (+/+) and Bβ2 KO (−/−) mice. D, Quantification of mitochondrial profile areas in neuronal somata and dendrites in cortex and hippocampus of the indicated genotypes. Plotted are means ± 95% CI of >60 somata and >20 dendrite segments per genotype from two mice of each genotype. ap < 0.05, bp < 0.01, ***p < 0.0001, ****p < 0.00005 compared with +/+ by ANOVA with Dunnett's post hoc test. Shown as Extended Data are cerebral blood flow during and after MCAO (Fig. 2-1), other mitochondrial parameters (Fig. 2-2), and expression of mitochondrial fission/fusion proteins (Fig. 2-3).
Figure 2-2
Mitochondrial morphology and content in forebrains of Bβ2−/− mice. A, Quantification of mitochondrial profile area from TEM images of the indicated brain regions. Plotted are means ± 95% CI of n = 49–80 neurons in amygdala and striatum and n = 29–34 hippocampal astrocytes from 2 mice/genotype. B, Quantification of mitochondrial volume density (% mitochondrial area/cytosolic area)/cell in the indicated brain regions. Plotted are mean ± 95% CI of n ≥ 60 neurons/condition from two mice/genotype. C, Western blots of mitochondrial proteins in forebrain homogenates of Bβ2+/+ and Bβ2−/− mice. D, Densitometry of immunoreactive bands in C relative to β-tubulin. Plotted are mean ± 95% CI of 6 mice for Bβ2+/+ and 4 mice for Bβ2−/−. Download Figure 2-2, TIF file.
Figure 2-3
Mitochondrial dynamics protein expression and phosphorylation in total forebrain homogenates. A, Western blot of the total and pS637 of the fission protein Drp1, as well as fusion protein optic atrophy 1 (Opa1) in forebrain homogenate. B, Quantification of blots in A relative to β-tubulin. Plotted are mean ± 95% CI of 6 mice for Bβ2+/+ and 4 mice for Bβ2−/−. Download Figure 2-3, TIF file.
Infarct volume was assessed 24 h following MCAO using a TTC method (Cheng et al., 1996). Animals were killed by isoflurane overdose, brains were rapidly removed, sectioned coronally at 2 mm intervals using a mouse Brain Matrix (Roboz Surgical Instrument) and immersed in TTC (2%) at 37°C for 20 min, followed by formaldehyde (4%) for 15 min. Sections were scanned, and infarct areas were measured using ImageJ (NIH) software. Scans of all brain slices were performed in order, such that any spurious, artefactual variations in dye intensity could be detected by reference to the previous and next slice. To correct for brain swelling due to edema after ischemia the corrected total infarct volume (%) was calculated according to: Corrected infarct volume (%) = {[volume of contralateral hemisphere − (volume of ipsilateral hemisphere-volume of infarct)]}/Volume of contralateral hemisphere × 100. MCAO surgery and infarct area determination were performed blinded, but with genotypes interleaved.
TEM of brain mitochondria
Six adult mice (2/genotype) were anesthetized with Nembutal (50 mg/kg) and transcardially perfused with PBS, pH 7.4, followed by 4% paraformaldehyde, and their brains removed. Samples were postfixed overnight in 2.5% glutaraldehyde, and 100 mm sodium cacodylate, pH 7.2, and cut at 100 mm using a vibratome before embedding flat in epoxy resin. Microscopy was performed blinded with a transmission electron microscope (JEOL 1230 TEM) equipped with a CCD camera. Analysis was performed using ImageJ to measure the size and shape of each mitochondrion blind to genotype.
Subcellular fractionation
Prior to forebrain isolation, mice anesthetized with ketamine/xylazine were perfused with an ice-cold PBS plus phosphatase inhibitor cocktail [0.1 mm ammonium molybdate, 2 mm EDTA, 2 mm EGTA, 50 mm NaF, 2 mm sodium pervanadate (made fresh), 10 mm sodium pyrophosphate, 50 nm Calyculin A, 2 μm FK506]. Forebrain tissue was isolated and homogenized in isolation buffer [IB; 225 mm mannitol, 75 mm sucrose, 1 mm EGTA, 1 mm EDTA, 5 mm HEPES-KOH, 2 mm sodium pervanadate (made fresh), 10 mm sodium pyrophosphate, 20 mm β-glycerophosphate, 1 mm benzamidine, leupeptin (1:2000), 100 nm microcystin-LR, 2 μm FK506, PMSF (1:250), pH 7.2] using a power homogenizer. One hundred microliters of homogenate sample was taken and diluted in 4× sample buffer.
A mitochondria-enriched fraction was then prepared similar to previously described (Wang et al., 2011). Briefly, the remaining homogenate was then centrifuged at 1100 × g for 2 min at 4°C keeping the supernatant [post-nuclear supernatant (PNS)] and resuspending the pellet in 4× sample buffer (nuclear fraction). A 100 µl sample of PNS was taken and diluted in 4× sample buffer. The remaining PNS was centrifuged at 17,000 × g for 15 min at 4°C. The supernatant from this step was further centrifuged at 80,000 × g for 60 min to obtain a cytosolic fraction while the pellet was resuspended in IB and centrifuged over a 9/10% Percoll interface at 18,500 × g for 15 min to obtain a mitochondria-enriched pellet. The mitochondria-enriched pellet was resuspended in IB and centrifuged at 10,000 × g for 5 min. The pellet was resuspended in 4× sample buffer as the mitochondrial fraction.
Stereotactic injection
Adult mice (12–14 weeks old) were anesthetized with 2–3% isoflurane and placed on a stereotaxic frame. Heat pads were used through the duration of the surgery to keep the body temperature stable. Eye ointment was applied to keep the eyes from drying. An incision was made to the skin to expose the skull after asepsis with Betadine and medical alcohol was applied. For viral injection, a craniotomy was made and a 1μL Hamilton syringe was slowly inserted into the target region. Virus (0.4–0.6 μl; 2.3 × 1013 vg/ml titer) was injected with an injection speed of 0.02 μl/min. For ex vivo Ca2+ imaging, AAV5-CAG-GCaMP6f-WPRE virus was unilaterally injected into the CA1 region of the hippocampus (AP: −1.95, ML: +1.25, DV: −1.2) of WT and Bβ2 KO mice. Animals were allowed to recover for 3 weeks after surgery before harvesting brains for acute slice preparation.
Acute slice preparation
Brains were rapidly dissected from mice following decapitation and placed in ice-cold N-methyl-D-glucamine (NMDG)-HEPES artificial CSF (aCSF) cutting solution (in mm): 92 NMDG, 2.5 KCl, 1.25 NaH2PO4, 30 NaHCO3, 20 HEPES, 25 glucose, 2 thiourea, 5 Na-ascorbate, 3 Na-pyruvate, 0.5 CaCl2·, and 10 MgSO4·. Brains were then sliced in 300 μm sections on a Leica VT-1000S vibratome (Leica Microsystems), placed in 34°C NMDG-HEPES aCSF cutting solution for 10–15 min, and then moved to 25°C recovery buffer (in mm: 92 NaCl, 2.5 KCl, 1.25 NaH2PO4, 30 NaHCO3, 20 HEPES, 25 glucose, 2 thiourea, 5 Na-ascorbate, 3 Na-pyruvate, 0.5 CaCl2·, and 10 MgSO4·) for at least 1 h before recording.
Two-photon Ca2+ imaging in slices
For Ca2+ imaging, slices were placed in a recording chamber on the stage of an upright Olympus FVMPE RS Multiphoton Microscope and imaged using λEx = 935 nm with a 20× water-immersion objective (NA = 1.0; Olympus, XLUMFLN) for changes in GCaMP6f fluorescence during perfusion with oxygenated (95% O2, 5% CO2) recording solution (in mm: 92 NaCl, 2.5 KCl, 1.25 NaH2PO4, 30 NaHCO3, 25 glucose, 2 CaCl2·, and 2 MgSO4·) for 5 min followed by perfusion with oxygen and glucose deprived (95% N2, 5% CO2) recording solution (in mm: 92 NaCl, 2.5 KCl, 1.25 NaH2PO4, 30 NaHCO3, 0 glucose, 25 sucrose, 2 CaCl2, and 2 MgSO4·) for 30 min. Image stacks were stabilized using the TurboReg plugin included with the Fiji distribution of ImageJ and fluorescence was quantified for the soma of individual cells following creation of a mask using the ImageJ multimeasure analysis function. For those neurons that produced delayed Ca2+ deregulation (DCD), latent period duration was measured as follows: the raw trace recording was smoothed using a running average protocol (GraphPad Prism 8); the first derivative to the smoothed trace was then calculated and plotted as a function of time; the times corresponding to onset of perfusion with oxygen and glucose deprived recording solution response and peak DCD slope were determined with the duration of time in minutes between the two calculated as the latent period duration.
Superoxide imaging in hippocampal slices
During incubation in recovery buffer at room temperature acute slices were incubated with 20 μm DHE for 1.5 h prior to imaging. For superoxide imaging slices were placed in a recording chamber on the stage of an Olympus FVMPE RS Multiphoton Microscope and imaged using λEx = 760 nm with a 20× water-immersion objective (NA = 1.0) for changes in DHE fluorescence during perfusion with oxygenated (95% O2, 5% CO2) recording solution (in mm: 92 NaCl, 2.5 KCl, 1.25 NaH2PO4, 30 NaHCO3, 25 glucose, 2 CaCl2·, and 2 MgSO4·; 0.02 DHE) followed by perfusion with oxygen and glucose deprived (95% N2, 5% CO2) recording solution (in mm: 92 NaCl, 2.5 KCl, 1.25 NaH2PO4, 30 NaHCO3, 0 glucose, 25 sucrose, 2 CaCl2·, and 2 MgSO4·; 0.02 DHE) for 5 and 30 min, respectively. Image stacks were aligned using the TurboReg plugin for Fiji/ImageJ and fluorescence was quantified for the soma of individual cells following creation of a mask using the ImageJ multimeasure analysis function. Rate of DHE oxidation (ΔFl/min, where ΔFl is change in DHE fluorescence intensity as a function of time) was calculated during constant perfusion of oxygen and glucose deprived recording solution.
Primary neuronal cultures
Primary mouse cultures of cortical and hippocampal neurons were prepared from P0-P1 neonates. Brains were removed from neonates and dissected first removing dura, dissecting hippocampi, and finally cortex. Tissue was stored in ice-cold Neurobasal Adult (NB-A; Invitrogen) media until processed. Tissue was incubated in HEPES buffered saline (HBS; 10 mm HEPES, 150 mm NaCl, pH 7.4, Ca2+/Mg2+-free), containing trypsin (0.03%) at 37°C for 20 min. The tissues were washed three times with Ca2+/Mg2+-free HBS before cells were dissociated by trituration. Cells were plated in NB-A complete [NB-A supplemented with B-27 (1.5×), glutamine (0.6 mm), and gentamycin (1:10,000)] plus 5% horse serum on either poly-l-lysine or poly-ornithine and laminin-coated plates or glass cover slips, respectively. Media was changed 4 h following plating to NB-A complete lacking serum. Half of the media volume was replaced with fresh NB-A complete every 4 d. Cells were maintained at 37°C in a humidified environment of 95% air/5% CO2.
Glutamate toxicity in primary neuronal cultures
At 14 DIV coverslips with cultured neurons were secured in a flow-through chamber fed HH buffer via a gravity perfusion system and mounted on an IX-71 epifluorescence microscope. For DCD and DHE experiments a baseline was established with a 2 min perfusion of HH buffer of the following composition: NaCl (140 mm), KCl (5 mm), CaCl2 (1.3 mm), MgCl2 (0.5 mm), MgSO4 (0.4 mm), KH2PO4 (0.4 mm), Na2HPO4 (0.6 mm), NaHCO3 (3 mm), HEPES (10 mm), d-glucose (10 mm), and at 310 mOsM, pH 7.4. The perfusion was then switched to a solution containing 100 μm glutamate, 10 μm glycine, and 200 nm TTX (Castilho et al., 1999).
Oxygen-glucose deprivation of neuronal cultures
For OGD, the medium of cortical cultures was exchanged on 14 DIV with NB-A medium without glucose (A24775, Invitrogen) that had previously been equilibrated with 94% N2, 5% CO2, and 1% O2. The cultures were maintained in a Billups-Rothenberg modular incubator chamber flushed in the same gas mixture for 30 min. After adding back the original, glucose-containing NB-A medium, cultures were returned to ambient oxygen and 5% CO2. Control cultures were mock-treated with normoglycemic/normoxic conditions. After 24 h, cultures were fixed in 4% paraformaldehyde and immunofluorescently labeled for MAP2B; nuclei were stained with Hoechst 33342 (1 μg/ml). Twenty epifluorescence images per culture were subjected to MAP2B densitometry and counts of nuclei using custom-written macros for ImageJ. Densitometry values and nuclei counts of OGD-treated cultures were normalized to mock-deprived cultures to determine percentage MAP2B and nuclei remaining.
Fura-FF cytosolic Ca2+ imaging
Cytosolic Ca2+ imaging was performed similarly to the protocol described by Schnizler et al. (2008). Cultures 14 DIV were incubated at 22°C for 30 min in 2 μm Fura-FF/AM (Invitrogen) and 0.01% pluronic acid in 2 ml HH buffer. The coverslip was then secured in a flow-through chamber fed HH buffer via a gravity perfusion system and mounted on an inverted IX-71 epifluorescence microscope (Olympus). Fura-FF fluorescence was sequentially excited at 340 nm and 380 nm via a Polychrome V monochromator (TILL Photonics) with either a 20× lens or a 40× oil-immersion objective lens (Olympus). Fluorescent emission at 510 (80) nm was collected at 0.2 Hz sampling frequency by a Photonics IMAGO CCD camera coupled to TillVisION live acquisition software. [Ca2+]i was quantified as the background subtracted fluorescence ratio (340/380 nm) multiplied by 100.
Analysis of DCD latency
For those neurons that produced DCD, latent period duration was measured as follows: the raw trace recording (converted to [Ca2+]i as described) was smoothed using a running average protocol (SigmaPlot 13 software); the first derivative to the smoothed trace was then calculated and plotted as a function of time; the times corresponding to peak glutamate [Ca2+]i response slope and peak DCD slope were determined; the duration of time in minutes between the two peaks of the slope was calculated as the latent period duration.
Superoxide imaging in primary neuronal cultures
At 14 DIV coverslips with cultured neurons were secured in a flow-through chamber fed HH buffer via a gravity perfusion system and mounted on an IX-71 epifluorescence microscope (Olympus). DHE (5 μm; Invitrogen) was perfused in with glutamate and DHE fluorescence was excited at 520 nm with emission collected at 605 nm with a 20× or 40× oil-immersion lens (Olympus). Rate of DHE oxidation (ΔF/min, where ΔF is change in DHE fluorescence intensity as a function of time) was calculated during constant perfusion of glutamate.
Respirometry
Oxygen consumption rate (OCR) was analyzed using a Seahorse Biosciences XF 96 Extracellular Flux analyzer (Agilent Technologies) essentially as described previously (Flippo et al., 2018). Briefly, primary cortical neurons were plated on 96-well Seahorse Biosciences plates coated with poly-l-lysine. At 14 DIV NB-A media was replaced with unbuffered DMEM (DMEM base medium supplemented with 10 mm glucose, 1 mm sodium pyruvate, 2 mm l-glutamine, pH 7.4) and incubated at 37°C for 1 h prior to respiration analysis. Four OCR readings were taken for each condition and inhibitors were injected in the following order; (1) basal, (2) oligomycin (2 μm), (3) FCCP (2 μm), and (4) rotenone/antimycin A (10 μm each). OCR was normalized to cell number following the assay by cresyl violet staining, measuring absorbance at 540 nm on a Biotek plate reader, and converting according to a set of standards of known cell number.
Experimental design and statistical analysis
Number of animals, brain sections, cultures, and cells analyzed for each experiment in this study are specified in the figure legends and were justified using power analysis. Approximately equal proportions of male and female mice were used for all experiments. Data were analyzed by Student's t test (two-tailed, with Welch's correction when appropriate) for single comparisons and by one-way ANOVA followed by Dunnett's multiple-comparisons test for multiple comparisons (GraphPad Prism v7 or v8 for Windows; Fig. 1-3). Unless noted otherwise, means ± 95% CI are plotted throughout, and significance levels are abbreviated as follows: *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001.
Results
Loss of Bβ2 has no apparent adverse consequences
We generated a global mouse knock-out of Bβ2 by replacing the first coding exon of the Ppp2r2b gene with lacZ and a polyadenylation sequence to terminate the transcript (Fig. 1-1A,B). Loss of Bβ2 was confirmed by qPCR (Fig. 1-1C) and by Western blotting PP2A holoenzymes purified by microcystin-Sepharose from total forebrain (Fig. 1-1D). Levels of the alternative promoter variant Bβ1, which is also brain-specific but localizes to the cytosol (Dagda et al., 2003), were unchanged in heterozygous and homozygous Bβ2 KO mice (Fig. 1-1D). Total levels of the four PP2A regulatory subunits belonging to the Ppp2r2 family (Bα/β/γ/δ, recognized by a pan-specific antibody) were also not detectably affected (Fig. 1-1D), suggesting that the Bβ2 KO removes a relatively small pool of PP2A holoenzymes from the brain. Bβ2 KO mice are born in Mendelian ratios, live a typical life span, and gain weight at the same rate as their WT littermates (Fig. 1A). Motor-coordination and motor-learning as assessed by the Rotarod test were likewise unaffected by deletion of the PP2A subunit (Fig. 1B). Anxiety levels as inferred from preference of the periphery in the open-field test was normal (Fig. 1C,D), as were activity measures, including distance traveled (Fig. 1-2A), velocity (Fig. 1-2B), and number of rearings in the open field (Fig. 1-2C). Diurnal activity levels were recorded over 23 h with a home cage monitoring system. Again, the Bβ2 KO did not impact day/night variations and cumulative measures of distance traveled (Fig. 1-2D), mobility (Fig. 1-2E), and time spent near the food hopper (Fig. 1-2E).
Bβ2 KO decreases infarct volume following MCAO and increases mitochondrial profile area in neurons
To evaluate whether deletion of Bβ2 influences cerebral ischemic sensitivity in vivo mice were subjected to transient (45 min) MCAO, which resulted in similar levels of blood flow reduction in Bβ2 KO and WT mice (Fig. 2-1). Indicative of improved cell survival, cerebral infarct volume was decreased by 40–60% in mice lacking one or both Bβ2 alleles (Fig. 2A,B). Examining mitochondrial ultrastructure revealed that heterozygous and homozygous Bβ2 deletion increased mitochondrial profile area in neurons of the hippocampus and cortex (Fig. 2C,D), but not of hippocampal astrocytes or neurons in the amygdala and striatum (Fig. 2-2A). No genotype differences in mitochondrial content were observed, consistent with a selective effect of Bβ2 deletion on the mitochondrial fission/fusion equilibrium as opposed to total mitochondrial mass (Fig. 2-2B,D). Additionally, we observed no changes in total protein expression of Drp1 or the mitochondrial fusion enzyme Opa1 (Fig. 2-3) indicating that altered mitochondrial morphology was not due to changes in expression of mitochondrial fission/and fusion genes.
Deletion of Bβ2 reduces PP2A association with mitochondria and promotes phosphorylation of Drp1 at S637
Given we observed no changes in total protein expression of mitochondrial fission/fusion enzymes to determine if changes in the mitochondrial fission/fusion equilibrium observed in mice lacking Bβ2 were related to Drp1-S637 phosphorylation and localization of PP2A to mitochondria we assessed levels of each in mitochondrial fractions isolated from the forebrain of Bβ2−/− and WT mice. In agreement with the observed increase in mitochondrial profile area, suggestive of inhibited mitochondrial fission, PP2A catalytic subunit levels were decreased, whereas Drp1-S637 phosphorylation was increased in mitochondrial fractions from Bβ2−/− forebrains (Fig. 3). In total forebrain homogenates, in contrast, Drp1 levels and phosphorylation were unaffected (Fig. 2-3), consistent with the notion that the majority of Drp1 is cytosolic and that PP2A/Bβ2 selectively regulates the mitochondria-associated pool of Drp1.
Deletion of Bβ2 reduces PP2A association with mitochondria and promotes phosphorylation of Drp1 at S637. A, Immunoblots for the PP2A catalytic subunit (PP2AC), total Drp1, and phosphorylated Drp1-S637 of mitochondrial fractions from Bβ2 WT (+/+) and Bβ2 KO (−/−) forebrains. B–D, Densitometric quantification of immunoblot signals relative to total protein (Ponceau S, PoncS; B, C) or total Drp1 (D). Shown are mean ± 95% CI of six WT and four Bβ2 KO mice. ***p < 0.001, ****p < 0.0001 by unpaired, two-tailed t test.
Deletion of Bβ2 delays calcium overload and blunts superoxide generation induced by OGD in the hippocampal CA1 region
Loss of Ca2+ homeostasis in neurons during ischemic insult, known as DCD, contributes to neuronal death during cerebral ischemia. To determine if deletion of Bβ2 improves the ability of neurons to maintain Ca2+ homeostasis during an ischemic insult, we expressed the fluorescent Ca2+ reporter GCAMP6f by stereotaxic delivery of virus into the hippocampus of WT and Bβ2 KO mice. To model the environment of the ischemic core, intracellular Ca2+ levels in response to OGD were measured by two-photon imaging of the CA1 region of acutely prepared hippocampal slices (Fig. 4A). In agreement with the in vivo cerebral ischemia results, deletion of Bβ2 prolonged the latency to DCD in CA1 neurons as measured as the time until the steepest rise in GCaMP6 fluorescence (Fig. 4B,C).
Deletion of Bβ2 delays calcium dysregulation and slows superoxide production in the CA1 region of the hippocampus during OGD. A, Representative two-photon microscopy image of GCaMP6 expression in the CA1 region of an acute hippocampal slice. B, Average traces of GCaMP6 fluorescence intensity before and during OGD (gray bar). C, The latency to DCD was determined as the time from the onset of OGD to the peak of the first derivative (rate) of fluorescence intensity for each genotype. D, Representative two-photon images of DHE fluorescence (pseudocolored according to the scale at the bottom; blue, lowest; red, highest) at rest and at the indicated times of OGD in the CA1 hippocampus. E, Average traces of DHE fluorescence intensity before and during OGD (gray bar). F, Superoxide (O2-) production rates were determined from linear fits of the DHE fluorescence traces during OGD. Mean ± 95% CI (B, gray and pink lines; E, gray areas) are plotted from 106 and 159 cells from three and four slices from 3 Bβ2+/+ mice and 3 Bβ2-/- mice, respectively, in B and C, and from 118 and 147 cells from four slices each from 3 Bβ2+/+ mice and 3 Bβ2-/- mice, respectively, in E and F. ****p < 0.0001 by Welch's two-tailed t test.
Mitochondrial reactive oxygen species, in particular superoxide play a prominent role in ischemic neuronal death, and bursts of superoxide generation may trigger DCD (Vesce et al., 2004). To monitor superoxide production, acute hippocampal slices were preloaded and then continuously superfused with DHE, a cell-permeable dye that fluoresces and intercalates into DNA on oxidation by superoxide. The Bβ2 KO slowed basal superoxide production from 6.6 ± 1.6 to 4.6 ± 0.6 ΔFl/min, resulting in lower baseline DHE fluorescence when two-photon imaging of the CA1 region commenced (Fig. 4D,E). Moreover, the KO all but eliminated the rapid rise in superoxide production that OGD triggered in hippocampal slices from WT mice(Fig. 4E,F).
Deletion of Bβ2 and Drp1 S637 phosphorylation delays calcium overload and inhibits superoxide production induced by glutamate excitotoxic challenge in primary hippocampal neurons
Although the results above indicated that genetic removal of Bβ2 can delay DCD and blunt toxic superoxide accumulation, it was unclear whether these effects were cell autonomous and whether they may also occur in areas adjacent to the ischemic core. During cerebral stroke, neurons in the ischemic core release excitotoxic levels of glutamate, which kills neurons in the surrounding ischemic penumbra via mitochondrial Ca2+ overload and/or DCD (Nicholls and Budd, 1998; Ramos-Cabrer et al., 2011). To determine whether the effects of Bβ2 deletion and additionally Drp1 S637 phosphorylation were cell autonomous in response to glutamate excitotoxicity we turned to using primary hippocampal neuron cultures from WT and Bβ2 KO mice. When challenged with 100 μm glutamate, cultured hippocampal Bβ2−/− neurons exhibited an increased latency to DCD compared with WT neurons (Fig. 5A–C), which was of similar magnitude as in hippocampal slices deprived of oxygen and glucose (compare Fig. 4B,C). Implicating increased Drp1-S637 phosphorylation in this neuroprotective effect, lentiviral replacement of endogenous with phosphomimetic Drp1 (GFP-Drp1 S637D; Flippo et al., 2018) in WT hippocampal neurons mimicked the effects of Bβ2 deletion (Fig. 5D–F). Consistent with DHE imaging in oxygen and glucose-deprived hippocampal slices, we found that Bβ2−/− neurons displayed lower rates of superoxide production under glutamate perfusion (Fig. 6A–C). Similar to the DCD experiments, phosphomimetic Drp1 substitution in WT neurons again mimicked, and perhaps even surpassed the detoxifying effects of deleting Bβ2 (Fig. 6D–F).
Deletion of Bβ2 and Drp1 S637 phosphorylation delays calcium overload induced by glutamate excitotoxic challenge. A, B, Individual (gray) and average (black, red) traces of cytosolic [Ca2+] in hippocampal neurons cultured from Bβ2+/+ and Bβ2−/− mice in response to 30 s pulses of 10 and 100 μm glutamate (arrows) and to sustained perfusion with 100 μm glutamate. Plotted are 340/380 excitation ratios times 100. C, Latency to DCD was calculated from traces in A and B as the time between the peaks of the first derivative (or rate) of the first and second [Ca2+]i rise following chronic glutamate treatment. D, E, Individual (gray) and average (black, red) traces of cytosolic [Ca2+] in response to 30 s pulses of 10 and 100 μm glutamate (arrows) and to sustained perfusion with 100 μm glutamate in WT hippocampal neurons following lentiviral transduction of either outer-mitochondrial (om)GFP or shRNA-mediated replacement of endogenous Drp1 with constitutively phosphorylated GFP-Drp1 (S637D). F, DCD latency in response to 100 μm glutamate was determined as in C. Note that 6 of 34 omGFP expressing neurons (18%), but 15 of 37 Drp1-S637D expressing neurons (41%) did not undergo DCD during the recording period. Data points and traces represent individual neurons cultured from three to four mice/condition with mean ± 95% CI plotted throughout. ***p < 0.001, ****p < 0.0001 by Welch's (C) and unpaired, two-tailed t tests (F).
Deletion of Bβ2 and Drp1 S637 phosphorylation inhibits superoxide production under excitotoxic glutamate challenge. A, Representative images show color-coded fluorescence intensity of the superoxide indicator DHE in hippocampal neurons (14 DIV) cultured from WT (+/+) and Bβ2 KO (−/−) mice at rest and at the indicated times after the start of 100 μm glutamate superfusion. The color lookup table scale is shown at the bottom. B, DHE fluorescence intensity traces (individual, gray; average, black or red) in response to 30 s (arrows) or constant (bar) glutamate superfusion in Bβ2+/+ and Bβ2-/- hippocampal neurons. C, Summary plots of superoxide production rates after glutamate treatment. D–F, superoxide production assays as described for A–C, but involving WT hippocampal cultures infected with lentivirus expressing omGFP or GFP-Drp1 S637D. Shown are representative images (D), traces (E), and rate quantification (F). Data points and traces represent individual neurons cultured from 3 to 4 mice/condition with mean ± 95% CI plotted throughout. ****p < 0.0001 by Welch's (C) and unpaired, two-tailed t tests (F).
Bβ2 deletion improves spare respiratory capacity and neuronal survival after ischemia
Ca2+ homeostasis fails when ischemia compromises mitochondrial ATP production (Nicholls et al., 2007; Yadava and Nicholls, 2007). Measuring oxygen consumption of cultured neurons subjected to mitochondrial stress tests, we found that spare respiratory capacity was increased both by deletion of the Drp1 phosphatase Bβ2 (Fig. 7A) and by phosphomimetic replacement of Drp1 (Fig. 7B). Further, deletion of Bβ2 protected against death of cultured neurons challenged by OGD, as measured by a decrease in neuronal MAP2B staining (Fig. 7C,D) and loss of nuclei (Fig. 7E). These results indicate that Bβ2 regulates ischemic sensitivity in a neuron-autonomous fashion. We recently reported that deletion of AKAP1, which recruits protein kinase A to mitochondria, exacerbates ischemic brain damage via loss of inhibitory Drp1 phosphorylation at Ser637 (Flippo et al., 2018). We crossed mice lacking Bβ2 with mice lacking AKAP1 and found that deletion of the Drp1 phosphatase normalized Drp1 S637 phosphorylation levels to WT levels (Fig. 7E). Coincident with restoration of inhibitory Drp1 phosphorylation, deletion of Bβ2 in mice lacking AKAP1 completely rescued the worsened stroke outcome associated with AKAP1 deletion (Fig. 7F), implicating Drp1 S637 as the common effector regulating ischemic sensitivity in vivo.
Bβ2 deletion improves spare respiratory capacity and neuronal survival after ischemia. A, Spare respiratory capacity (SRC) determined as the area-under-the curve (AUC) of uncoupled respiration minus basal respiration measured in cortical neurons cultured from WT (+/+) and Bβ2 KO (−/−) mice. B, Spare respiratory capacity of cortical neurons expressing omGFP or constitutively phosphorylated GFP-Drp1 (S637D). C, D, Survival of cortical cultures from +/+ and −/− mice subjected to 45 min of OGD was assessed 24 h later by immunofluorescence labeling of MAP2B, a protein that is rapidly degraded in challenged neurons (C, representative images; D, densitometry), and by automated counts of nuclei stained by Hoechst 33 342 (E, densitometry and counts normalized to normoxic conditions). F, Forebrain mitochondrial fractions from three mice either WT for or lacking AKAP1 and Bβ2 were assayed for Drp1 phosphorylation by immunoblotting. G, Mice of the indicated genotypes were subjected to 30 min MCAO and infarct volumes were determined after 24 h by TTC staining as in Figure 2, A and B. Data for WT and AKAP1 KO mice were previously reported (Flippo et al., 2018) and are shown here for comparison to double KO mice. For A, B, and G data points represent mice, with n = 12 for Bβ2+/+, 8 for Bβ2-/-, 9 for omGFP, 6 for Drp1-S637D, 13 for AKAP1+/+/Bβ2+/+, and 6 each for AKAP1-/-/Bβ2+/+ and AKAP1-/-/Bβ2-/-. D, E, Data points represent individual cultures from four WT and five Bβ2-/- mice. Mean ± 95% CI are plotted throughout. *p < 0.05, **p < 0.01, ***p < 0.005, for 7A-E -/- was compared with +/+ using unpaired, two-tailed t tests and in G one-way ANOVA was used for comparison between genotypes. Quantitation of electron transport chain complexes is shown in Figure 7-1.
Figure 7-1
Expression of ETC complex subunits in forebrain homogenates. A, Western blot for subunits of the indicated ETC complexes in forebrain homogenate (OXPHOS antibody mixture, Abcam). B, Quantification of blot in A relative to total protein (Ponc S, Ponceau S) staining. Plotted are mean ± 95% CI of 6 mice for Bβ2+/+ and 4 mice for Bβ2−/−. *p < 0.05 by multiple t tests. Download Figure 7-1, TIF file.
Discussion
Here, we identified a role for the neuronal and mitochondria-targeted PP2A/Bβ2 holoenzyme in opposition to the neuroprotective AKAP1/PKA complex, acting on an evolutionary conserved inhibitory phosphorylation site on Drp1. Previous preclinical studies suggested that pharmacological Drp1 inhibition could have therapeutic benefit limiting brain damage after ischemic stroke (Grohm et al., 2012; Zhao et al., 2014; Li et al., 2015). However, the specificity and efficacy of the Drp1 inhibitor mdivi-1 has been challenged (Bordt et al., 2017). Importantly, our work here maintains a critical role for Drp1-mediated mitochondrial fission in dictating sensitivity to ischemic injury and suggests that phosphorylation of Drp1 at S637 functions as an important moderator of stroke outcomes.
Drp1 is also phosphorylated at S616, and cyclin-dependent kinases, extracellular signal-regulated kinases, and protein kinase C have been implicated as S616 kinases. However, in contrast to S637, phosphorylation at S616 was reported to activate Drp1, promoting mitochondrial fission (Taguchi et al., 2007; Qi et al., 2011; Strack et al., 2013; Kashatus et al., 2015). We detected no change in Drp1 S616 phosphorylation levels in total lysates or mitochondrial fractions from Bβ2 KO brains (data not shown). This suggests that either PP2A/Bβ2 does not dephosphorylate S616, or that kinases dominate over phosphatases at this site. High basal stoichiometry of Drp1 S616 phosphorylation (∼40%; Strack et al., 2013) argues for the latter explanation.
Ischemia and excitotoxicity trigger rapid mitochondrial fragmentation; however, this is thought to be independent of Drp1 and instead caused by osmotic swelling of mitochondria (Barsoum et al., 2006; Shalbuyeva et al., 2006; Brustovetsky et al., 2009; Young et al., 2010; Slupe et al., 2013). In this context, PP2A/Bβ2 and the opposing AKAP1/PKA likely influence stroke outcomes not by promoting and inhibiting mitochondrial fragmentation during the ischemic onslaught, but by shaping the mitochondrial network prior to the insult, thereby determining the likelihood that neurons survive in the ischemic penumbra. Indeed, the AKAP1 KO reduces (Flippo et al., 2018), whereas the Bβ2 KO increases mitochondrial spare respiratory capacity (Fig. 7A), an indicator of resilience to metabolic stress (Nicholls et al., 2007; Yadava and Nicholls, 2007). The mechanistic relationship underlying mitochondrial elongation and improved metabolic capacity deserves further study.
Although we did not attempt to quantify whether Bβ2 deletion rescues some brain regions within the territory of the middle cerebral artery better than others from stroke, such regional selectivity would indeed be predicted. Indeed, according to electron microscopy, the Bβ2 KO leads to mitochondrial elongation in the cortex and hippocampus (Fig. 2D), but not in the striatum and amygdala (Fig. 2-2A). Higher Bβ2 expression in the former compared with the latter regions is one possible explanation. Additionally, available single-cell RNA sequencing data http://mousebrain.org/) suggest that Drp1 itself is more highly expressed in the cortex and hippocampus than the striatum, which may increase the effect magnitude of the Bβ2 KO on mitochondrial shape in the cortex and hippocampus relative to the striatum.
Although we generated a global KO of Bβ2, we did not detect any adverse behavioral or physiological consequences in the homozygous KO. This may in part be rationalized by nervous system-specific expression of the Bβ gene (Dagda et al., 2003; one of 12 PP2A regulatory subunit genes) and maintained expression of the more abundant Bβ1 splice variant in the Bβ2 KO brain (Fig. 1-1B). Still, we observed robust mitochondrial elongation in neurons and reduction of focal stroke damage even in heterozygous Bβ2 KO mice. Heterozygous effects are significant, in that they suggest that even partial reduction of PP2A/Bβ2 activity may be therapeutically relevant. However, our results also raise questions regarding the physiological function of this outer-mitochondrial PP2A holoenzyme. We previously reported that mitochondrial shape, PP2A/Bβ2, and AKAP1/PKA influence neurite outgrowth and synapse formation in vitro (Dickey and Strack, 2011). It will be important to assess if and how the Bβ2 KO impacts brain development in vivo and to challenge mice with additional behavioral, including cognitive tasks.
We found that displacing PP2A from mitochondria by deletion of its targeting subunit Bβ2 or by mimicking Drp1 S637 phosphorylation improves bioenergetic capacity of mitochondria to maintain Ca2+ homeostasis when hippocampal neurons are challenged by either OGD or glutamate excitotoxicity. Particularly striking was the robust inhibition of superoxide production that we observed in hippocampal neurons lacking Bβ2 ex vivo and in vitro. Superoxide toxicity is well established not only in acute ischemic neuronal injury, but also during the reperfusion phase, when mitochondrial ATP production resumes (Traystman et al., 1991; Chouchani et al., 2016). Our ex vivo and in vitro studies have only examined acute injury responses, so it is possible that Bβ2 influences later stages of stroke damage as well. It will be instructive to examine whether Bβ2 deletion (and other means of Drp1 inhibition) also blunts the superoxide surge during reperfusion, because this would extend the therapeutic window of targeting mitochondrial fission.
Although overall mitochondrial mass is unaffected (Fig. 2-2), we did detect increased expression of electron transport chain complex subunits in forebrain lysates of Bβ2−/− mice (Fig. 7-1), providing a possible explanation for increased spare respiratory capacity in Bβ2 KO neurons. Increased ETC complex expression and electron flux may also boost the antioxidant defense system, for instance in the form of homeostatic increases in superoxide scavengers and detoxifying enzymes that could occur in Bβ2 KO neurons. Although further work is required to establish the precise mechanism by which mitochondrial shape impacts bioenergetic capacity and the ability to limit superoxide toxicity during an ischemic insult, our work establishes the Drp1 S637 phosphorylation state as a pivotal regulator of neuronal susceptibility to ischemic stroke.
Footnotes
This work was supported by National Institutes of Health Grants NS056244, NS087908 to S.S., NS087068 and NS096246 to Y.M.U., and HL139926, NS109910 to A.K.C.; by American Heart Association Grant 18EIA33900009 to A.K.C., by the Roy J. Carver Charitable Trust and Iowa Neuroscience Institute (S.S., Y.M.U.), and by these facilities of the Carver College of Medicine: Genomics Division, Viral Vector Core, and Central Microscopy Core.
The authors declare no competing financial interests.
- Correspondence should be addressed to Stefan Strack at stefan-strack{at}uiowa.edu or Yuriy M. Usachev at yuriy-usachev{at}uiowa.edu
