The Ca2+-gated channel TMEM16A amplifies capillary pericyte contraction and reduces cerebral blood flow after ischemia

Pericyte-mediated capillary constriction decreases cerebral blood flow in stroke after an occluded artery is unblocked. The determinants of pericyte tone are poorly understood. We show that a small rise in cytoplasmic Ca2+ concentration ([Ca2+]i) in pericytes activated chloride efflux through the Ca2+-gated anion channel TMEM16A, thus depolarizing the cell and opening voltage-gated calcium channels. This mechanism strongly amplified the pericyte [Ca2+]i rise and capillary constriction evoked by contractile agonists and ischemia. In a rodent stroke model, TMEM16A inhibition slowed the ischemia-evoked pericyte [Ca2+]i rise, capillary constriction, and pericyte death; reduced neutrophil stalling; and improved cerebrovascular reperfusion. Genetic analysis implicated altered TMEM16A expression in poor patient recovery from ischemic stroke. Thus, pericyte TMEM16A is a crucial regulator of cerebral capillary function and a potential therapeutic target for stroke and possibly other disorders of impaired microvascular flow, such as Alzheimer’s disease and vascular dementia.

using acute cortical slices or aged mice ( 9) or by cardiac perfusion under terminal anesthesia for in vivo experiments . Animal breeding, experimental procedure and methods of killing were conducted in accordance with the UK Home Office regulations (Guidance on the Operation of Animals, Scientific Procedures Act, 1986).

Human tissue
Human cortical tissue, removed to gain access to underlying tumours in patients, was provided by Dr Huma Sethi, a neurosurgeon at the National Hospital for Neurology and Neurosurgery, Queen Square. The tissue was obtained from five patients: three glioblastoma patients (a 50-year-old male, a 66-year-old male and a 74-year-old female), one thymus cancer patient (a 40-year-old female) and one sarcoma metastasis patient (a 46-year-old female). All work was performed with the informed consent of the patients and ethical approval from the National Health Service (REC number 15/NW/0568 and IRAS project ID 180727).

Human genetics
Genetic association summary data for TMEM16A expression were obtained from the Genotype Tissue Expression (GTEx) project (gtexportal.org). The version 8 dataset, which included 15,201 samples across 49 tissues from a multi-ethnic group of 838 individuals, was used to identify single-nucleotide polymorphism genetic variants in the TMEM16A gene that associated with altered TMEM16A expression in any of the tissues considered in the database.
Expression data from GTEx version 7, including 10,294 samples across 48 tissues from a multi-ethnic group of 620 individuals, were used to examine the association of variants with pooled TMEM16A expression across all tissues. Genome-wide association study (GWAS) summary data for diastolic blood pressure were obtained from a study of 757,601 European ancestry participants (4). GWAS summary data for risk of developing ischemic stroke and recovery after ischemic stroke were obtained from the MEGASTROKE and GISCOME consortia, respectively. The MEGASTROKE study included 60,341 ischaemic stroke cases and 454,450 controls (5). The GISCOME study included 6,021 stroke cases, adjusted for baseline stroke severity (as measured by the National Institutes of Health Stroke Scale), and measured outcome 60-190 days after ischemic stroke using the modified Rankin Scale (a commonly used score for measuring disability after stroke, ranging from no symptoms to death (6)). The GISCOME study analysis consisted of 3,741 patients that had a score of 0-2 (more favourable outcome) and 2,280 patients with a score of 3-6 (less favourable outcome).
Full details of all these studies have been published previously (GTEx Consortium, 2017) (4)(5)(6). All studies from which data were used for this work had already obtained relevant authorisation, and further ethical approval or participant consent was not required. The identified genetic proxy for TMEM16A activity, the single nucleotide polymorphism rs755016, was further investigated in a phenome-wide association study using the Medical Research Council Integrative Epidemiology Unit Open GWAS Project (7,8) (https://gwas.mrcieu.ac.uk/, accessed 17 November 2021).
Cortical pericytes on the 1 st , 2 nd and 3 rd capillary branch order of penetrating arterioles (PAs) were imaged by acquiring Z-stacks (within 30-100 μm cortical depth, 2 µm step size, 150 to 300 nm pixel size, 1.58 to 6.3 µs pixel dwell time) every minute using a 20 x / 1.0 NA water immersion objective (W Plan-Apochromat, Zeiss). Emitted fluorescence was spectrally divided by a 555-nm dichroic mirror and collected by GaAsP detectors. The power under the objective did not exceed 20 mW. Image processing was performed in FIJI (11). To account for movements in the x-and y-dimensions, image stacks were projected at maximum intensity in the z-dimension and co-registered using the StackReg plugin in FIJI. Changes in Patch-clamping (detailed below) and brightfield imaging of capillary pericytes in rat cortical slices was performed at 30-50 μm depth within cortical layers III and IV, and images were recorded using an Olympus BW51 microscope equipped with differential interference contrast (DIC), a 40x water immersion objective, a Coolsnap HQ2 or an IRIS 9TM Scientific CMOS camera CCD camera, and ImagePro Plus (Media Cybernetics) or Metafluor (Molecular Devices) acquisition softwares. Images were acquired every 5 or 30 s, with an exposure time of 50 ms. The pixel size was 160 nm. During imaging, any changes in focus were manually restored by carefully adjusting the focus knob between image acquisitions.
Internal capillary diameters were measured by manually placing a measurement line perpendicular to the capillary at pericyte somata (Figure 2A, C) using MetaMorph software.

Ischaemia pericyte death experiments
Cortical slices were prepared from Sprague Dawley rats as described above. Slices were transferred into conical flasks under OGD or control (aCSF) conditions in the absence of presence of Ani 9 (2 M). Solution were gassed with 5% CO2 and 95% N2 (for OGD solutions) or 5% CO2 and 95% O2 (for aCSF solutions) at 37 °C for the entire duration of the experiment. The solutions were supplemented with 7.5 μM of the necrosis marker propidium iodide (PI; Sigma-Aldrich, #81845) and kept in the dark for the entire duration of the experiment. After the 1 hr incubation, the slices were swiftly washed 3 times (using the above incubation solutions in the absence of PI), prior to transfer to 12-well plates for fixation in 4% paraformaldehyde for 1 hr on a rocking shaker. Following 3 washes in PBS, slices were incubated in 10 µg/mL IB4 in blocking buffer (10% (vol/vol) horse serum, 0.3% (vol/vol) Triton X-100, 1.5% (wt/vol) glycine and 1% (wt/vol) bovine serum albumin in PBS) overnight at 4°C on a shaker. Slices were then washed 3 times in PBS before incubation with the nuclear stain DAPI in PBS (1:50,000) on a shaker at room temperature for 1 hr.

Electrophysiology
Whole-cell patch-clamp recordings were performed on an Olympus BW51 microscope (as for brightfield imaging) using a HEKA EPC9 amplifier, controlled via the built-in analogue-to-digital and digital-to-analogue converter and the Patchmaster version 2x32 software (HEKA Elektronik, Lambrecht, Germany). Pipettes were pulled from borosilicate glass capillary tubes (Harvard Apparatus, UK) using a Narishige PC-10 pipette puller (Narishige, Japan). Pipette tip diameter yielded a resistance of 4-5 MΩ in the working solutions. The series resistance was compensated to achieve a maximal effective series resistance lower than 10 MΩ. Currents were filtered at 2 kHz and sampled at 10 kHz. Liquid junction potentials (LJPs) were calculated (12,13) and corrected for off-line; these LJPs equalled ~-9 and -13 mV for Cs + and K + based solutions (see below), respectively. Families of TMEM16A-CaCC currents (Figure 1 C, D) were elicited in response to 1 s pulses from -100 mV to +100 mV in 10 mV increments, each followed by a 0.5 s step to -60 mV and elicited every 3 s from a holding potential of -40 mV; with all command voltages adjusted for the LJP of -9 mV before display. The whole-cell currents shown in Figure 4E were recorded in response to 2 s voltage ramps from -100 to +100 mV from a holding potential of -40 mV, with all command voltages adjusted for the LJP of -13 mV before display.
Rat cortical slices were prepared as described above and incubated at ~34°C for 20

In vivo ischaemia two-photon imaging, hypoxia measures and TTC staining
NG2-Cre ERT2 -GCaMP5G mice (P32-P63), NG2-dsRed mice (P30-P83) or aged wildtype mice (15 months) were anesthetised using urethane (1.55 g/kg given in two doses 15 minutes apart; Sigma-Aldrich, #94300) and anesthesia was confirmed by the lack of a withdrawal reflex to a paw pinch. Body temperature was maintained at 36-37°C using a feedback-controlled heating pad. Eyes were protected from drying by applying polyacrylic acid eye drops (Dr. Winzer Pharma GmbH). The trachea was cannulated and mice were mechanically ventilated with medical air supplemented with oxygen using a MiniVent (Model 845). Sutures were placed around the left and right common carotid arteries. The skull was exposed, slightly thinned using a drill and dried using compressed air. A custommade headplate was centred over the right barrel cortex, 3 mm laterally from the midline and immediately caudal to the coronal suture. The headplate was attached using superglue gel and mice were head fixed to a custom-built stage. The skull was thinned over the left frontal cortex to attach 1 or 2 laser Doppler flowmetry probes spaced ~ 5mm apart. Cerebral blood flow (CBF) was measured with an OxyFlo Pro laser Doppler system. Laser Doppler traces were extracted in Matlab 2015b. A craniotomy of approximately 2 mm diameter was performed and the dura was removed. The exposed barrel cortex was superfused for 1 hr with 10 µM Ani9 or aCSF vehicle in HEPES-based aCSF and sealed off with 2% agarose in HEPES-buffered aCSF (containing Ani9 or vehicle) covered by a glass coverslip. All mice in the Ani9 and vehicle groups were age-matched to minimise variability in leptomeningeal anastomoses that occurs with aging (14). Neutrophil staining using the Alexa Fluor 647 anti-mouse Ly-6G rat monoclonal antibody (2.5 µg/ml in blocking buffer, BioLegend, #127610) was performed on 100 μm thick brain sections (following an overnight incubation in blocking buffer) from NG2-Cre ERT2 -GCaMP5G mice or NG2-dsRed mice that underwent cardiac FITC-albumin perfusion at ~1.5 hrs after common carotid artery occlusion. Following 4 x 10 min washes in PBS, brain slices were mounted with DAPI for confocal imaging. The shortest distance from the periphery of an intravascular neutrophil to the centre of a pericyte was measured using 3D capillary tracing with the Simple Neurite Tracer plugin in FIJI.

Reagent and drug preparation
Stock concentrations of drugs dissolved in DMSO were 50 mM for Ani9 and MONNA and 100 mM for bumetanide. ET-1 (Sigma-Aldrich, E7764) stock was 100 µM in H2O and U46619 (Cayman, #16450) stock was 28.5 mM in methyl acetate. The final concentration of DMSO or methyl acetate in aCSF solutions never exceeded 0.1% v/v or 0.0007% v/v, respectively. Ani9 required brief warming under continuous stirring to be dissolved in aCSF. The photosensitive L-type voltage-gated Ca 2+ channel blocker nimodipine was prepared (protected from light) fresh from powder in polyethylene glycol 400 at 4.78 mM (PEG-400; Sigma-Aldrich, #202398). Nimodipine was then further diluted in 15% (2hydroxypropyl)-β-cyclodextrin (Sigma-Aldrich, H107) in PBS. Urethane stock was prepared in saline at 0.18 g/ml. Tamoxifen stock was dissolved overnight in corn oil at 10 mg/ml on a shaker at room temperature before oral gavage.

Genetic association analysis
All supporting data for this work are available within the article text, Table 1 and   Supplementary Table 1. Statistical analysis was performed using the programme R (version 3.4.2). The genetic proxy for TMEM16A activity was selected as a single-nucleotide polymorphism in the TMEM16A gene that: (i) associated with increased expression of TMEM16A in any tissue at genome-wide significance after applying a Bonferroni correction for multiple testing of 48 tissues (p<10 -9 ), and (ii) had a secondary association with increased diastolic blood pressure (P<0.05). Clumping was performed to a pairwise linkage disequilibrium threshold of r 2 <0.001 using the TwoSampleMR package of R (8). Elevated diastolic blood pressure was used as a secondary trait by which to select the genetic proxy for TMEM16A activity due to the established role of the TMEM16A channel in controlling the tone and diameter of peripheral resistance arteries, which are a major determinant of diastolic blood pressure (17). The association of variants with pooled TMEM16A expression across all tissues was estimated using fixed-effects meta-analysis (18). For Mendelian randomization, the association of the variant proxying TMEM16A activity with (i) risk of ischemic stroke and (ii) recovery after ischemic stroke were examined. A statistical significance threshold of p<0.025 was used, after applying a Bonferroni correction for testing of two outcomes. For exploration of pleiotropy, the PhenoScanner database of genetic associations was searched to explore for potential pleiotropic effects of the identified genetic proxy for TMEM16A activity that could bias Mendelian randomization analysis (19). Given that PhenoScanner mostly contains summary data from GWAS analyses, a genome-wide significance threshold was used for identifying such pleiotropic associations (P<5x10 -8 ).

Group sizes and exclusion criteria
Group size for each specific experiment were equal by design and informed by power calculation conducted at priori. No exclusion criteria were applied with the exception of: (i) in vivo CCAO procedure: experiments in which the mean CBF during CCAO did not drop by >85% of the baseline average were excluded in accord with Grubbs' method (20) and (ii) ex vivo electrophysiological studies: cells used for patch-clamp experiments had seal resistance ≥ 1 G at the start of the experiment. If resistance was lower than this value when the wholecell configurations was achieved, the experiment was terminated.

Randomisation and Blinding
Mice or rats of appropriate age were selected randomly on the day of the experiment.
For ex vivo experiments, pericytes (in capillaries of a given order), that were imaged or patch-clamped were randomly selected in isolated brain slices.

8B
Cumulative probability of block as a function of distance from pericyte CCAO, n=110, N=10