Piezo1 suppression reduces demyelination after intracerebral hemorrhage

Abstract

Piezo1 is a mechanically-gated calcium channel. Recent studies have shown that Piezo1, a mechanically-gated calcium channel, can attenuate both psychosine- and lipopolysaccharide-induced demyelination. Because oligodendrocyte damage and demyelination occur in intracerebral hemorrhage, in this study, we investigated the role of Piezo1 in intracerebral hemorrhage. We established a mouse model of cerebral hemorrhage by injecting autologous blood into the right basal ganglia and found that Piezo1 was largely expressed soon (within 48 hours) after intracerebral hemorrhage, primarily in oligodendrocytes. Intraperitoneal injection of Dooku1 to inhibit Piezo1 resulted in marked alleviation of brain edema, myelin sheath loss, and degeneration in injured tissue, a substantial reduction in oligodendrocyte apoptosis, and a significant improvement in neurological function. In addition, we found that Dooku1-mediated Piezo1 suppression reduced intracellular endoplasmic reticulum stress and cell apoptosis through the PERK-ATF4-CHOP and inositol-requiring enzyme 1 signaling pathway. These findings suggest that Piezo1 is a potential therapeutic target for intracerebral hemorrhage, as its suppression reduces intracellular endoplasmic reticulum stress and cell apoptosis and protects the myelin sheath, thereby improving neuronal function after intracerebral hemorrhage.

Introduction

There are two main types of stroke: ischemic stroke, in which vessels in the brain are obstructed, and hemorrhagic stroke, in which blood leaks into the brain as a result of vascular rupture (Donnan et al., 2008). Intracerebral hemorrhage (ICH) is an extremely serious form of stroke that is impossible to treat (Keep et al., 2012). In Asia, approximately 20–30% of stroke cases are attributable to ICH (Ikram et al., 2012). It mostly occurs in the basal ganglia, a white matter-rich area that is especially vulnerable to mechanical stress in response to hemorrhage, resulting in hemiplegia, partial sensory impairment, ectopic blindness, and other complications (Jiang et al., 2019; Fu et al., 2021). White matter mainly consists of axon bundles and many kinds of glial cells like microglia, astrocytes, and oligodendrocytes. The main functions of oligodendrocytes are to wrap axons, forming an insulated myelin sheath structure that assists in the efficient transmission of bioelectrical signals (Boccazzi et al., 2022). However, the complex mechanisms underlying oligodendrocyte injury and demyelination post-ICH are not well understood. The lack of research in this area makes it challenging to develop therapeutic strategies to alleviate white matter injury in patients with ICH (Tao et al., 2017; Zuo et al., 2017).

The pathophysiology of ICH involves the primary injury, caused by hematoma-induced mechanical stress, and the secondary injury, caused by blood catabolites. ICH-induced white matter injury results in severe neurological dysfunction, which affects patient prognosis and quality of life. White matter injury often occurs within 3 days of ICH, so there is a relatively long therapeutic window in which to administer treatment (Wasserman and Schlichter, 2008). While there have been many studies of the pathophysiology of secondary injury from hematotoxic products after ICH, few studies have assessed the direct tissue destruction induced by hematoma-induced mechanical stress, also known as “mass effect” (Qureshi et al., 2009). There is controversy regarding the efficacy of surgically removing intracerebral hematomas, but the relationship between white matter injury and hematoma-induced mass effect deserves to be investigated. A previous study performed in a rat model of mechanical microballoon showed that hematoma-induced mass effect can directly cause white matter injury and that some mechano-sensitive channels might be involved in the pathophysiological process of this phenomenon (Ma et al., 2011).

Piezo1 is an extensively expressed mechanically gated calcium channel (Gnanasambandam et al., 2015, 2017; Yang et al., 2022). In bone biology, Piezo1 has been demonstrated to contribute to osteoclastogenesis induced by mechanical stress in the periodontal ligament cells of humans (Li et al., 2019). Furthermore, it has been demonstrated that Piezo1 contributes to the integration of vascular structures and physiological force (Li et al., 2014). Various cells in the central nervous system express Piezo1, including neurons, microglia, astrocytes, and endothelial cells. Recently, Piezo1 has also been shown to be expressed in oligodendrocyte precursor cells and to regulate their activity in the aging central nervous system (Segel et al., 2019; Velasco-Estevez et al., 2022). Recent evidence also suggests that inhibition of Piezo1 can attenuate both psychosine- and lipopolysaccharide-induced demyelination by blocking superfluous Ca²⁺ influx into neuronal axons, which may suppress calpain-mediated destabilization of integrin attachments, thereby promoting myelin formation (Velasco-Estevez et al., 2020). However, the role of Piezo1 suppression in mechanical stress-induced white matter injury and the development of neurological deficits after ICH is unclear.

Myelin basic protein (MBP), which is synthesized by oligodendrocytes, plays an important role in myelination by promoting the formation of the major dense line (Barbarese and Pfeiffer, 1981), and MBP breakdown is a marker of demyelination. Studies have shown that MBP breakdown may be caused by oligodendrocyte apoptosis and might contribute to demyelination after ICH, as well as to neuronal apoptosis (Zhuo et al., 2016; Fu et al., 2021; Darbinian and Selzer, 2022). Apoptosis can be initiated by a variety of pathways, including the death receptor pathway, the mitochondrial pathway, and the endoplasmic reticulum (ER) pathway (Orrenius et al., 2011; Mohammed Thangameeran et al., 2020). When ER homeostasis is disrupted, the cell suffers a series of biochemical changes, which include protein folding dysfunction, impaired protein transport, and disruption of calcium homeostasis (Groenendyk et al., 2021). When cells experience ER stress for a prolonged period of time, apoptosis-promoting factors are activated, leading to apoptotic cell death (Benavides et al., 2005). Previous studies have shown that the ER pathway can mediate oligodendrocyte apoptosis after ICH (Zhuo et al., 2016; Mohammed Thangameeran et al., 2020). Therefore, we speculated that Piezo1 may trigger apoptosis in oligodendrocytes following ICH by influencing ER Ca²⁺ homeostasis.

In the present study, we investigated the following hypotheses: (1) suppression of Piezo1 with a pharmacological antagonist ameliorates functional deficits and demyelination in a mouse model of ICH; (2) Piezo1 suppression improves myelination partly through reducing oligodendrocyte apoptosis after ICH; and (3) Piezo1 activation causes the disruption of Ca²⁺ homeostasis and irreversible ER stress, resulting in oligodendrocyte apoptosis following ICH injury.

Methods

Experimental animals

A total of 282 male adult C57BL/6 mice (aged 8–10 weeks, weighing 22–30 g, Experimental Animal Center of the Chinese PLA General Hospital, license No. SCXK (Jing) 2021-0006) were used in this study. All experiments were performed in strict accordance with the requirements specified in the Guide for the Care and Use of Laboratory Animals (8^th ed) (National Institutes of Health, 2011). Animal use protocols were approved by the Animal Ethic Committee of the Chinese PLA General Hospital (approval No. 2021-X13-04) on January 4, 2021. The mice were housed under specific pathogen-free conditions at a constant temperature (22°C) and humidity level (45%), with a 12/12-hour light/dark cycle and adequate food and water. There were 6–8 mice housed in each cage.

Only male mice were used in our study for the following reasons: first, estrogen fluctuates with the menstrual cycle and affects behavior; second, estrogen has a neuroprotective effect (Petrovska et al., 2012); and third, behavioral bias could be caused by different gonadal hormones.

Experimental design

The study involved three experiments.

Experiment I

To determine Piezo1 expression at different time points following ICH injury, 60 mice were randomly assigned to eight groups (Figure 1): sham (n = 12), ICH 3-hour (n = 12), ICH 6-hour (n = 6), ICH 12-hour (n = 6), ICH 24-hour (n = 6), ICH 48-hour (n = 6), ICH 72-hour (n = 6), and ICH 7-day (n = 6). Piezo1 levels were detected by western blot assay (n = 6 per group). The cellular location of Piezo1 was evaluated by double immunofluorescence staining of Piezo1 and oligodendrocyte transcription factor 2 (olig2, an oligodendrocyte marker) at 3 hours after ICH (n = 6 per group). Three mice were excluded (one mouse died from an anesthetic accident, and two mice died after surgery), and the excluded mice were replaced with new ones (Table 1).

Figure 1.

Experimental timelines.

ELISA: Enzyme-linked immunosorbent assay; IF: immunofluorescence; TEM: transmission electron microscopy; WB: western blotting.

Table 1.

The sample size in each experiment

	Group	IF	TEM	WB	Elisa	Behavioral tests	Brain water content 24 h	Brain water content 72 h	Excluded
Experiment I	Sham	6		6
	3 h	6		6
	6 h			6
	12 h			6
	24 h			6					1
	48 h			6
	72 h			6					1
	7 d			6					1
Experiment II	Sham					7	6	6	2
	ICH+Vehicle					7	6	6	4
	ICH+Dooku1-low					7	6	6	3
	ICH+Dooku1-mid					7	6	6	3
	ICH+Dooku1-high					7	6	6	2
Experiment III	Sham		6	6	6				2
	Sham+Dooku1-mid	6	6	6	6				5
	ICH+Vehicle	6	6	6	6				4
	ICH+Dooku1-mid	6	6	6	6				3

Elisa: Enzyme-linked immunosorbent assay; IF: immunofluorescence; TEM: transmission electron microscopy; WB: western blotting.

Experiment II

To determine the effective dose of the selective Pizeo1 antagonist Dooku1 (Evans et al., 2018), 95 adult mice were randomly divided into five groups (Figure 1): sham (n = 19), ICH + vehicle (n = 19), ICH + Dooku1-low (n = 19), ICH + Dooku1-mid (n = 19), and ICH + Dooku1-high (n = 19). Dooku1 was dissolved in dimethyl sulfoxide (Sigma-Aldrich, St. Louis, MO, USA) to a final concentration of less than 0.1%. For the 72-hour study, Dooku1 (5, 10, 20 mg/kg) was intraperitoneally administered four times, at 0.5, 12, 24, and 48 hours after operation. Neurological deficits were evaluated by the modified Garcia scale, beam balance test, and corner turn test at 24 and 72 hours after ICH (n = 7 per group). Brain water content was assessed at 24 and 72 hours after ICH in all groups (n = 6 per group for each time point). Fourteen mice were excluded (six mice died from an anesthetic accident, and eight mice died after surgery), and the excluded mice were replaced with new ones (Table 1).

Experiment III

The purpose of this experiment was to investigate the function of Dooku1 in myelin degradation after ICH, and its possible molecular mechanisms. Ninety-six adult mice were randomly divided into four groups (Figure 1): sham (n = 24), sham + Dooku1-mid (n = 24), ICH + vehicle (n = 24), and ICH + Dooku1-mid (n = 24). Myelin degradation was elevated, as determined by immunofluorescence staining of MBP and DMBP (n = 6), transmission electron microscopy (n = 6), and western blot assay for MBP expression (n = 6). Immunofluorescence staining was performed to detect the expression of ER stress–related protein glucose regulated protein 78 (GRP78) and C/EBP homologous protein (CHOP) in each group (n = 6). TdT-mediated dUTP nick end labeling and western blot assays for Bax and Bcl-2 were used to evaluate oligodendrocyte apoptosis (n = 6). Western blot assay was performed to detect the expression of phosphorylated inositol-requiring enzyme 1 (P-IRE1), inositol-requiring enzyme 1 (IRE1), caspase-12, caspase-3, phosphorylated protein kinase RNA-like endoplasmic reticulum kinase (P-PERK), protein kinase RNA-like endoplasmic reticulum kinase (PERK), activating transcription factor 4 (ATF4), phosphorylated eukaryotic initiation factor 2 alpha (P-eIF2α), and eukaryotic initiation factor 2 alpha (eIF2α) (n = 6). Enzyme-linked immunosorbent assay was performed to detect the expression of inflammation factors (interleukin-1 beta (IL-1β), interleukin-6 (IL-6), and tumor necrosis factor alpha (TNFα)) (n = 6). Fourteen mice were excluded (five mice died from an anesthetic accident, and nine mice died after surgery), and the excluded mice were replaced with new ones (Table 1).

The mice were anesthetized by intraperitoneal injection of 1% pelltobarbitalum natricum (50 mg/kg, Sigma-Aldrich) before modeling or sampling.

ICH modeling

The mice underwent autologous blood infusion to simulate ICH, as we described previously (Yang et al., 2018). The mice were fixed on a three-dimensional frame structure (RWD Life Sciences Ltd., Shenzhen, China) in a prone position. A hole was drilled into the parietal bone on the right side of the skull (2.0 mm lateral to the midline and 0.8 mm anterior to the bregma) (Yang et al., 2018). The tail tip was cut off, and blood was collected from the wound and quickly transferred to a 50-μL glass syringe. A sterile Hamilton syringe needle was used to inject 25 μL of autologous blood into the right basal ganglia at a rate of 2 μL/min using a microinfusion pump (KD Scientific, Holliston, MA, USA). Animals in the sham group were subjected to the same procedures as the ICH mice, but no blood was injected.

Behavioral assessments

Neurological scores were assessed at 24 and 72 hours after ICH as described below.

Beam balance test

The beam balance test is used to test the balancing ability of mice (Chen et al., 2020). Each mouse was placed alone in the center of the wooden beam. The distance and times that the mouse traveled in 1 minute were recorded, and a score of 0–4 points was assigned based on these values. The average score was calculated for three consecutive trials. Higher scores indicated better test performance.

Modified Garcia scale

The modified Garcia scale is an 18-point scoring system that is used to evaluate neurological deficits after ICH (Qu et al., 2018). It covers six different test areas, including spontaneous activity, body symmetry, front paw extension, climbing symmetry, body proprioception, and response to whisker stimulation. A maximum of three points can be earned for each test area, and higher scores indicate better test performance.

Corner turn test

The corner turn test is used to evaluate motor deficits. Each mouse was allowed to enter the test apparatus at an angle of 30°. To leave the corner, the mouse must turn to the left or right, and that decision is recorded. Each mouse was tested 10 times in total, with a 30-second interval between each iteration, and the percentage of right turns was calculated.

Evaluation of brain water content

Brain edema was evaluated using the wet-weight/dry-weight method (Zou et al., 2021). Mice were first anesthetized and then sacrificed at 24 or 72 hours after the operation. The brain tissue was then isolated to determine the wet weight. The brain tissue was then processed at 100°C for 24 hours and weighed to determine the dry weight. The water content of the brain was calculated using the formula: (wet weight – dry weight)/wet weight × 100%.

Western blotting

Twenty-four hours after the operation, the mice were anesthetized and the brain tissue surrounding the hematoma in the ipsilateral basal ganglia (around the lesion site) was collected using an inverted microscope (RWD Life Science). An extraction kit (Beyotime Biotechnology, Co., Ltd., Jiangsu, China) supplemented with a protease inhibitor cocktail (Roche, Indianapolis, IN, USA) was used to extract total protein from the brain tissue. Equal amounts of protein were loaded onto sodium dodecyl sulfate polyacrylamide gels, subjected to gel electrophoresis, and then transferred onto polyvinylidene fluoride membranes (Bio-Rad, Hercules, CA, USA, Cat# 1620184). The membranes were blocked in 3% bovine serum albumin for 2 hours, then incubated with the primary antibodies at 4°C overnight. The following primary antibodies were used: rabbit polyclonal anti-Pizeo1 antibody (1:500, Alomone Labs, Jerusalem, Israel, Cat# APC-087), mouse monoclonal anti-MBP antibody (1:1000, Santa Cruz Biotechnology, Santa Cruz, CA, USA, Cat# sc-271524, RRID: AB_10655672), rabbit polyclonal anti-CHOP antibody (1:500, Sigma-Aldrich, Cat# SAB4500631, RRID: AB_10743042), rabbit polyclonal anti-GRP78 antibody (1:1000, Abcam, Cambridge, UK, Cat# ab21685, RRID: AB_2119834), rabbit monoclonal anti-Bax antibody (1:2000, Abcam, Cat# ab182733), rabbit monoclonal anti-P-IRE1 antibody (1:1000, Abcam, Cat# ab124945, RRID: AB_11001365), rabbit polyclonal anti-IRE1 antibody (1:1000, Abcam, Cat# ab37073, RRID: AB_775780), rat monoclonal anti-caspase12 antibody (1:500, Sigma-Aldrich, Cat# C7611, RRID: AB_476869), rabbit monoclonal anti-caspaase3 antibody (1:3000, Abcam, Cat# ab32351, RRID: AB_725946), rabbit monoclonal anti-cleaved caspase-3 antibody (1:2000, Abcam, Cat# ab214430) rabbit monoclonal anti-P-PERK antibody (1:1000, Cell Signaling Technology, Danvers, MA, USA, Cat# 3179s), rabbit monoclonal PERK antibody (1:1000, Cell Signaling Technology, Cat# 3192, RRID: AB_2095847), rabbit monoclonal ATF-4 antibody (1:1000, Cell Signaling Technology, Cat# 11815, RRID: AB_2616025), rabbit monoclonal anti-phospho-eukaryotic initiation factor 2 alpha (eIF2α) (Ser51) antibody (1:1000, Cell Signaling Technology Cat No: 3597, RRID: AB_390740), rabbit polyclonal anti-eIF2α antibody (1:1000, Abcam, Cat# ab26197, RRID: AB_2096478), and rabbit monoclonal anti-Bcl-2 antibody (1:1000, Abcam, Cat# ab182858, RRID: AB_2715467). Rabbit polyclonal anti-β-actin antibody (1:1000, Proteintech, Wuhan, China, Cat# 81115-1-RR) rabbit polyclonal anti-β-tublin antibody (1:1000, Proteintech, Cat# 11224-1-AP, RRID: AB_2210206), and rabbit polyclonal anti-glyceraldehyde-3-phosphate dehydrogenase (GAPDH) antibody (1:1000, Proteintech, Cat# 10494-1-AP, RRID: AB_2263076) were used as internal loading controls. The membranes were then incubated with appropriate secondary antibodies and incubated at 37°C for 1 hour. The following secondary antibodies were used: horseradish peroxidase (HRP)-conjugated Affinipure rabbit anti-goat IgG (H+L) (1:2000, Proteintech, Cat# SA00001-4, RRID: AB_2864335), HRP-conjugated Affinipure goat anti-rabbit IgG (H+L) (1:2000, Proteintech, Cat# SA00001-2, RRID: AB_2722564), HRP-conjugated Affinipure goat anti-rat IgG (H+L) (1:2000, Proteintech, Cat# SA00001-15, RRID: AB_2864369), and HRP-conjugated Affinipure goat anti-mouse IgG (H+L) (1:2000, Proteintech, Cat# SA00001-1, AB_2722565). Enhanced chemiluminescence (Amersham Biosciences, Arlington Heights, IL, USA) was used to visualize the immunoreactive bands, were observed using an imaging system (VersaDoc MP 4000; Bio-Rad). Image Lad software (version 5.2.1, National Institutes of Health, Bethesda, MD, USA) was used to analyze the results. Uncropped western blot images are shown in Additional Figure 1 ^{(2.4MB, tif)} .

Terminal-deoxynucleoitidyl transferase mediated nick end labeling assay

At 24 hours after ICH, a commercial terminal-deoxynucleoitidyl transferase mediated nick end labeling (TUNEL) assay kit (FITC; In Situ Cell Death Detection Kit, Roche Molecular Biochemicals, Mannheim, Germany) was used to detect apoptosis. Whole brain slices were incubated with 20 mg/mL protease K dissolved in 10 mM Tris(hydroxymethyl)aminomethane hydrochloride for 15 minutes at 37°C. Brain tissue sections were then incubated with 0.3% H₂O₂ solution prepared with anhydrous methanol to block endogenous peroxidase. After 10 minutes, the sections were rinsed with phosphate-buffered saline (PBS). The brain slices were then placed in a solution of 0.1% sodium citrate and 0.1% TritonX-100 and incubated at 4°C for 2 minutes. After rinsing with PBS, each brain tissue slice was incubated with 50 μL TUNEL reaction solution for 1 hour at 37°C in the dark. After several washes with PBS, the nuclei were counterstained with 4′,6-diamidino-2-phenylindole (DAPI) for 5 minutes. The TUNEL-positive apoptotic cells were detected by fluorescence emission at 488 nm.

Transmission electron microscopy

To examine axon microstructure, transmission electron microscopy was performed at 24 hours after ICH. Anesthetized mice were transcardially infused with 4% paraformaldehyde (in PBS) 24 hours after ICH induction. Tissue from the ipsilateral basal ganglia (around the lesion site) was fixed for 15 minutes with glutaraldehyde and formaldehyde (2%). After fixation was complete, the tissue block was trimmed into a 1-cm³ cube and stored at 4°C overnight. The next day, the samples were dehydrated, sliced, and stained with lead citrate and uranyl acetate. Then, the tissue slice components and surface structure were observed using a transmission electron microscope (Hitachi, Tokyo, Japan). The white matter (WM) content was determined by calculating the G-ratio (axon diameter/axon + myelin diameter), as we described previously (Qu et al., 2018).

Immunofluorescence staining

Immunofluorescence staining was used to examine axon morphology and the expression of ER stress-related proteins. Anesthetized mice were transcardially perfused with 4% paraformaldehyde (in PBS) at 24 hours after ICH induction. The frozen sections (20-mm thick) from the ipsilateral basal ganglia were heated at 37°C for 30 minutes and then washed with 0.01 M PBS. After incubation for 30 minutes in 5% bovine serum albumin, the sections were incubated with primary antibodies overnight at 4°C. The antibodies used were as follows: rabbit polyclonal anti-Pizeo1 antibody (1:100, Alomone Labs, Cat# APC-087, RRID: AB_2756743), mouse monoclonal anti-Olig2 antibody (1:100, Sigma-Aldrich, Cat# SAB5300407), mouse anti-MBP antibody (1:100; Santa Cruz, Cat# sc-271524, RRID: AB_10655672), rabbit anti-degraded MBP (dMBP) antibody (1:250; Millipore, St. Louis, MO, USA, Cat# AB5864), rabbit polyclonal anti-Chop antibody (1:100, Sigma-Aldrich, Cat# SAB4500631, RRID: AB_10743042), and rabbit polyclonal anti-GRP78 antibody (1:100, Abcam, Cat# ab21685, RRID: AB_2119834). After washing in phosphate-buffered saline three times, the sections were incubated with appropriate secondary antibodies at 26 ± 1°C for 2 hours. The secondary antibodies used were as follows: donkey anti-mouse IgG (H+L) ReadyProbes™ Alexa Fluor™ 594 antibody (1:200, Invitrogen, Cat# R37115 RRID: AB_2556543), donkey anti-goat IgG (H+L) highly cross-adsorbed secondary antibody, Alexa Fluor™ Plus 555 (1:200, Invitrogen, Cat# A32816, RRID: AB_2762839), donkey anti-rabbit IgG (H+L) highly cross-adsorbed secondary antibody, and Alexa Fluor™ 488 antibody (1:500, Invitrogen, Cat# A-21206, RRID: AB_2535792). The positive cells fluoresced at specific wavelengths that were observed and photographed using a confocal microscope (Zeiss, LSM780).

Enzyme-linked immunosorbent assay

To evaluate the degree of inflammation, enzyme-linked immunosorbent assay (ELISA) was performed to detect inflammatory factors (IL-1β, IL-6, and TNF-α) in all four groups. At 24 hours after ICH, the mice were anesthetized, and the brain tissue surrounding the hematoma was collected using an inverted microscope (RWD Life Science). ELISA assay kits (Beyotime, Shanghai, China, Cat# PI301, PI326, PT512) were used to detect the expression of inflammatory factors (IL-1β, IL-6, and TNF-α) according to the manufacturer‘s instructions. The concentration of each protein was calculated based on absorbance at 450 nm.

Statistical analysis

The sample sizes were determined based on those used in a previous study (Qu et al., 2018). The evaluator was blinded to the group assignments. All data are expressed as the mean ± standard deviation (SD) and were analyzed using GraphPad Prism 6.01 (GraphPad Software, San Diego, CA, USA, www.graphpad.com). One-way analysis of variance followed by Scheffe’s post hoc test was employed for comparisons among three and more groups. P < 0.05 was considered statistically significant.

Results

Time course of Piezo1 expression in mice after ICH

Piezo1 protein levels were detected by western blotting to explore the changes in Piezo1 expression following ICH in experiment I (Figure 2A). Compared with the sham group, Piezo1 expression was significantly higher 3 to 48 hours after ICH (P < 0.05) and peaked at 3 hours (Figure 2B and C). After the 48-hour time point, the Piezo1 expression declined rapidly, and within 72 hours of the cerebral hemorrhage, the symptoms of the mice in the ICH group were similar to those seen in the sham group (P > 0.05; Figure 2B and C). Immunofluorescence staining performed 3 hours after ICH demonstrated that Piezo1-positive cells also expressed Olig2, an oligodendrocyte marker (Joseph et al., 2016; Figure 2D).

Figure 2.

Time course and location of Piezo1 expression following ICH.

(A) Injury region and sampling sites after ICH in mice. (B, C) Representative western blot image and quantification of Piezo1 expression in sham and ICH mice at 3, 6, 12, 24, 48, and 72 hours and 7 days after ICH. (D) Representative images of immunofluorescence staining for Piezo1 (green-Alexa Fluor 488) and Olig2 (red-Alexa fluor Plus 555) 3 hours after ICH. Scale bars: 50 μm (upper row), 20 μm (lower row). Data are expressed as mean ± SD (n = 6 per time point). *P < 0.05, vs. sham group (one-way analysis of variance followed by Scheffe’s post hoc test). DAPI: 4′,6-Diamidino-2-phenylindole; ICH: intracerebral hemorrhage; Olig2: oligodendrocyte lineage transcription factor 2.

Piezo1 suppression attenuates brain edema and neurological deficits after ICH

In experiment II, the Piezo1 antagonist Dooku1 (Evans et al., 2018) was used to explore the possible role of Piezo1 following ICH. Three different concentrations (5, 10, and 20 mg/kg) were used to explore the effectiveness of systemic Dooku1 administration. Behavioral tests were performed 24 and 72 hours after ICH to evaluate the effects of Dooku1 on neurological function. No toxic effects of Dooku1 were observed in the modified Garcia test, the corner turn test, or the beam balance test.

ICH mice showed significantly greater neurobehavioral deficits than sham mice in all tests (all P < 0.05; Figure 3A–C). In the balance beam test, treatment with 10 mg/kg Dooku1 significantly improved scores 24 hours after ICH (P < 0.05, vs. ICH + vehicle group; Figure 3A). Treatment with the other two concentrations (5 and 20 mg/kg) also improved scores compared with the ICH + vehicle group, but the differences were not significant (P > 0.05, vs. ICH + vehicle group; Figure 3A). In the modified Garcia test, treatment with 5 mg/kg Dooku1 did not affect the score 24 and 72 hours after ICH (both P > 0.05, vs. ICH + vehicle group; Figure 3B), treatment with 10 and 20 mg/kg Dooku1 significantly improved the score at 72 hours after ICH, but only 10 mg/kg Dooku1 significantly improved the score 24 hours after ICH (P < 0.05, vs. ICH + vehicle group; Figure 3B). In the corner turn test, only treatment with 10 mg/kg Dooku1 significantly increased the right turn percentage 72 hours after ICH (P < 0.05, vs. ICH + vehicle group; Figure 3C). Furthermore, there was a noticeable rise in brain water content in the ICH mice at 24 and 72 hours after ICH. Compared with the ICH + vehicle group, only 10 mg/kg Dooku1 treatment significantly decreased the brain water content at 24 and 72 hours after ICH (P < 0.05, vs. ICH + Vehicle group; Figure 3D).

Figure 3.

The effects of Dooku1 on neurological deficits and brain edema following ICH.

(A) Balance beam test score, (B) modified Garcia test score, and (C) corner turn test score at 24 and 72 hours after ICH. (D) Brain water content at 24 and 72 hours after ICH. Data are expressed as mean ± SD (n = 7 for each group in the balance beam test, modified Garcia test, and corner turn test; n = 6 per group for brain water content). *P < 0.05, vs. sham group; #P < 0.05, vs. ICH + vehicle group (one-way analysis of variance followed by Scheffe’s post hoc test). ICH: Intracerebral hemorrhage.

Piezo1 suppression alleviates myelin loss and disorganization after ICH

Approximately 30% of central nervous system myelin is made up of MBP, which is used as a protein marker of demyelination following ICH (Joseph et al., 2016). In experiment III, immunofluorescence staining showed myelinating clusters red or green complete circles. Fewer MBP-positive myelinating clusters were observed after ICH injury than in the sham mice (P < 0.05, vs. sham group; Figure 4A), and treatment with Dooku1 (10 mg/kg) significantly increased the number of MBP-positive myelinating clusters at 24 hours after ICH (P < 0.05, vs. ICH + vehicle group, Figure 4A and B). dMBP is produced by myelin degradation (Xia et al., 2019) and is used as a biomarker of demyelination. At 24 hours after ICH, the number of dMBP-positive myelinating clusters was significantly increased (P < 0.05, vs. sham group; Figure 4A and C), and treatment with Dooku1 (10 mg/kg) markedly decreased the number of dMBP-positive myelinating clusters (P < 0.05, vs. ICH + vehicle group; Figure 4A and C). Western blot analysis indicated that MBP expression levels were also significantly lower at 24 hours after ICH (P < 0.05, vs. sham group; Figure 4E and F). Treatment with Dooku1 (10 mg/kg) markedly increased MBP expression (P < 0.05, vs. ICH + vehicle group; Figure 4E and F).

Figure 4.

The effects of Dooku1 on white matter injury following ICH.

(A) Representative immunofluorescence staining for MBP (red-Alexa Fluor 594) and DMBP (green-Alexa Fluor 488) and representative ultrastructural changes in the myelin sheath as detected by transmission electron microscopy 24 hours after ICH. The density of MBP-positive myelinating clusters in the ICH + Dooku1-mid group was higher than that in ICH + vehicle group. In contrast, the density of DMBP-positive myelinating clusters in the ICH + Dooku1-mid group was lower than that in ICH + vehicle group. The transmission electron microscopy results showed that the myelin sheath layers were thin and unsecured in the ICH + vehicle group, while they were tight and regular in the ICH + Dooku1-mid group. Scale bars: 100 μm (first two rows); 2 μm (third row). (B–D) Quantitative analysis of MBP-positive myelinating clusters (B), DMBP-positive myelinating clusters (C), and G-ratio around the ipsilateral basal ganglia (D). The images used to generate these graphs are provided in Additional file 1 ^{(91.9KB, pdf)} . (E, F) Representative western blot images and quantification of MBP expression in each group at 24 hours after ICH. Data are expressed as mean ± SD (n = 6 per group). *P < 0.05, vs. sham group; #P < 0.05, vs. ICH + vehicle group (one-way analysis of variance followed by Scheffe’s post hoc test). DMBP: Degraded myelin basic protein; ICH: intracerebral hemorrhage; MBP: myelin basic protein.

Next, transmission electron microscopy was used to observe the ultrastructure of myelinated nerve fibers in the ipsilateral basal ganglia following ICH. The myelin sheath layers were tight and regular in the sham group. At 24 hours after ICH, the myelin sheath layers were thin and unsecured, leading to a significant increase in the G-ratio (P < 0.05, vs. sham group; Figure 4A and D). However, treatment with Dooku1 (10 mg/kg) alleviated the degree of demyelination and significantly decreased the G-ratio (P < 0.05, vs. ICH + vehicle group; Figure 4A and D).

Piezo1 suppression alleviates oligodendrocyte apoptosis after ICH

In experiment III, the role of Piezo1 in oligodendrocyte apoptosis in the ipsilateral basal ganglia region was assessed by administering Dooku1 (10 mg/kg) and assessing cell apoptosis by TUNEL assay 24 hours after ICH. At 24 hours after ICH, the number of TUNEL-positive oligodendrocytes was significantly greater in the ICH groups than in the sham group (P < 0.05, vs. sham group; Figure 5A and B), while the number of TUNEL-positive oligodendrocytes in the ICH + Dooku1-mid group was lower than that in the ICH + vehicle group (P < 0.05, vs. ICH + vehicle group; Figure 5A and B).

Figure 5.

The effects of Dooku1 on oligodendrocyte apoptosis following ICH.

(A) Representative images of double immunofluorescence staining for Olig2 (red-Alexa Fluor Plus 555) and TUNEL (FITC) in the ipsilateral basal ganglia (around the lesion sites) 24 hours after ICH. At 24 hours after ICH, the density of TUNEL⁺ Olig2⁺ cells in the ICH + Dooku1-mid group was less than that in the ICH + vehicle group. Scale bars: 200 μm (left four columns), 40 μm (right column). (B) Quantitative analyses of the ratio of TUNEL⁺ Olig2⁺/Olig2⁺ cells around the ipsilateral basal ganglia. The images used to generate these graphs are provided in Additional file 1 ^{(91.9KB, pdf)} . (C–F) Representative western blot images and quantitative analyses of Bax (C, D) and Bcl-2 (E, F) expression 24 hours after ICH. Data are expressed as mean ± SD (n = 6 per group). *P < 0.05, vs. sham group; #P < 0.05, vs. ICH + vehicle group (one-way analysis of variance followed by Scheffe’s post hoc test). DAPI: 4′,6-Diamidino-2-phenylindole; ICH: intracerebral hemorrhage; Olig2: oligodendrocyte lineage transcription factor 2; TUNEL: TdT-mediated dUTP nick-end labeling.

Bax and Bcl-2 are both key factors in cell apoptosis (Edlich, 2018). Bax expression was significantly increased and Bcl-2 expression was significantly decreased around the hematoma in the ipsilateral basal ganglia after ICH (both P < 0.05, vs. sham group; Figure 5C–F), while treatment with Dooku1 (10 mg/kg) significantly upregulated Bax expression and downregulated Bcl-2 expression (both P < 0.05, vs. ICH + vehicle group; Figure 5C–F).

Piezo1 suppression attenuates ICH-induced ER stress in oligodendrocytes (Experiment III)

GRP78 is an important indicator of ER stress (Sano and Reed, 2013). In experiment III, at 24 hours after ICH, the number of GRP78-positive, olig2-positive cells was significantly higher in the ICH groups than in the sham group (P < 0.05). Compared with the ICH + vehicle group, the number of GRP78-positive oligodendrocytes in the ICH + Dooku1-mid group was significantly lower (P < 0.05; Figure 6A and B). The western blot results also showed significant elevation of GRP78 expression 24 hours after ICH compared with the sham group (P < 0.05). Treatment with Dooku1 (10 mg/kg) significantly decreased GRP78 expression (P < 0.05, vs. ICH + vehicle group; Figure 6C and D).

Figure 6.

The effects of Dooku1 on GRP78 expression in oligodendrocytes following ICH.

(A) Representative images of double immunofluorescence staining for GRP78 (green-Alexa Fluor 488) and Olig2 (red-Alexa Fluor Plus 555) in the ipsilateral basal ganglia (around the lesion sites) 24 hours after ICH. At 24 hours after ICH, the density of GRP78⁺ olig2⁺ cells in the ICH + Dooku1-mid group was less than that in the ICH + vehicle group. Scale bars: 200 μm (left four columns), 40 μm (right column). (B) Quantitative analysis of the ratio of GRP78⁺ Olig2⁺/Olig2⁺ cells around the ipsilateral basal ganglia. The images used to generate these graphs are provided in Additional file 1 ^{(91.9KB, pdf)} . (C, D) Representative western blot images and quantification of GRP78 expression in each group 24 hours after ICH. Data are expressed as mean ± SD (n = 6 per group). *P < 0.05, vs. sham group; #P < 0.05, vs. ICH + vehicle group (one-way analysis of variance followed by Scheffe’s post hoc test). GRP78: Glucose regulated protein 78; ICH: intracerebral hemorrhage; Olig2: oligodendrocyte lineage transcription factor 2.

CHOP is also important for ER stress-induced apoptosis (Sano and Reed, 2013) and is known to be involved in ICH-induced apoptosis (Liew et al., 2019). At 24 hours after ICH, the number of CHOP-positive, olig2-positive cells was significantly higher in the ICH groups compared with the sham group (P < 0.05). Compared with the ICH + vehicle group, the number of GRP78-positive oligodendrocytes in the ICH + Dooku1-mid group was significantly lower (P < 0.05; Figures 7A and B). The western blot results also showed a significant increase in CHOP expression at 24 hours after ICH (P < 0.05) compared with the sham group. Treatment with Dooku1 (10 mg/kg) significantly decreased CHOP expression (P < 0.05, vs. ICH + vehicle group; Figure 7C and D).

Figure 7.

The effects of Dooku1 on CHOP expression in oligodendrocytes following ICH.

(A) Representative images of double immunofluorescence staining for CHOP (green-Alexa Fluor 488) and Olig2 (red-Alexa Fluor Plus 555) in the ipsilateral basal ganglia (around the lesion sites) 24 hours after ICH. The density of CHOP⁺ Olig2⁺ cells in the ICH + Dooku1-mid group was less than that in the ICH + vehicle group. Scale bars: 200 μm (left four columns), 40 μm (right column). (B) Quantitative analyses of the ratio of CHOP⁺ Olig2⁺/Olig2⁺ cells around the ipsilateral basal ganglia. The images used to generate these graphs are provided in Additional File 1 ^{(91.9KB, pdf)} . (C, D) Representative western blot images and quantification of GRP78 expression 24 hours after ICH. Data are expressed as mean ± SD (n = 6 per group). *P < 0.05, vs. sham group; #P < 0.05, vs. ICH + vehicle group (one-way analysis of variance followed by Scheffe’s post hoc test). CHOP: C/EBP homologous protein; ICH: intracerebral hemorrhage; Olig2: oligodendrocyte lineage transcription factor 2.

To further investigate the role of Dooku1 in ER stress and ER stress-related apoptosis after ICH, we measured the expression of ER stress-related proteins was measured. A previous study demonstrated that IRE1α can activate downstream apoptosis-related proteins, particularly caspase-12, which causes caspase-3 activation, leading to apoptosis (Fischer et al., 2002). We found that p-IRE1α expression increased significantly after ICH, while IRE1α expression did not change markedly. Therefore, the p-IRE1α/IRE1α ratio was significantly higher (P < 0.05; Figure 8A and B) compared with the sham group. Treatment with Dooku1 (10 mg/kg) decreased p-IRE1α expression, but IRE1α expression was unaffected compared with the ICH + vehicle group (P < 0.05; Figure 8A and B). Caspase-12, caspase-3, and cleaved-caspase-3 expression showed similar changes to that of p-IRE1α (all P < 0.05; Figure 8A and B). Taken together, these findings show that Dooku1 can inhibit the IRE1α-caspase-12-caspase-3 signaling pathway and function as a protective factor after ICH.

Figure 8.

The effects of Dooku1 on ER stress-related protein expression in the ipsilateral basal ganglia (around the lesion sites) 24 hours after ICH.

(A, B) Representative western blot images and quantification of p-IRE1, IRE1, caspase-12, caspase-3, and cleaved caspase-3 expression. (C, D) Representative western blot images and quantification of p-PERK, PERK, ATF4, p-eIF2α, and Eif2α expression (vs. β-tubulin). Data are expressed as mean ± SD (n = 6 per group). *P < 0.05, vs. sham group; #P < 0.05, vs. ICH + vehicle group (one-way analysis of variance followed by Scheffe’s post hoc test). ATF4: Activating transcription factor 4; ER: endoplasmic reticulum; ICH: intracerebral hemorrhage; IF2α: eukaryotic initiation factor 2 alpha; IRE1: inositol-requiring enzyme 1; P-eIF2α: phospho-eukaryotic initiation factor 2 alpha; PERK: protein kinase RNA-like endoplasmic reticulum kinase; P-IRE1: inositol-requiring enzyme 1; P-IRE1: phosphor Inositol-requiring enzyme 1; P-PERK: phosphor-protein kinase RNA-like endoplasmic reticulum kinase.

Under ER stress, PERK is activated and phosphorylates elF2α. The phosphorylated form of eIF2α then activates ATF4, which upregulates CHOP expression, eventually leading to apoptosis (Harding et al., 2003). Western blot analysis demonstrated a significant elevation in p-PERK expression 24 hours after ICH, while PERK expression did not change significantly. Therefore, the p-PERK/PERK ratio increased significantly compared with the sham group (P < 0.05; Figure 8C and D). Treatment with Dooku1 (10 mg/kg) significantly reduced the p-PERK/PERK ratio at 24 hours after ICH compared with the ICH + vehicle group (P < 0.05; Figure 8C and D). The p-eIF2α/eIF2α ratio was also higher 24 hours after ICH, and Dooku1 (10 mg/kg) significantly decreased the p-eIF2α/eIF2α ratio (both P < 0.05; Figure 8C and D). ATF4 expression changed in a similar pattern to the p-eIF2α/eIF2α ratio (both P < 0.05; Figure 8C and D). Taken together, these findings showed that Piezo1 suppression relieved ICH-induced ER stress by inhibiting the eIF2α/PERK/ATF4 pathway.

Discussion

In the present study, we explored the effect of the pressure-activated ion channel Piezo1 on demyelination following ICH. Our studies showed that Piezo1 expression significantly increased at 3 hours after ICH injury, that a moderate dose (10 mg/kg per mouse) of the Piezo1 antagonist Dooku1 relieved neurological impairment and myelin damage, and that this effect was likely related to a reduction in oligodendrocyte apoptosis. As such, Piezo1 suppression may be a novel therapeutic strategy for treating patients with ICH.

As the most important cellular component of white matter, oligodendrocytes are susceptible to hemorrhagic impairment (Kang and Yao, 2019). According to previous studies, oligodendrocyte apoptosis contributes to demyelination after ICH, accompanied by proliferation of oligodendrocyte precursor cells in the ipsilateral basal ganglia during the acute period (Zhu et al., 2012; Joseph et al., 2016; Zhuo et al., 2016). All oligodendrocyte-lineage cells express the transcription factor Olig2, and our double immunofluorescence staining experiments showed that cells in the ipsilateral basal ganglia expressed both that Olig2 and Piezo1 after ICH. Moreover, the western blotting results revealed that Piezo1 expression significantly increased at 3 hours after ICH injury. Due to the sensitivity of Piezo1 to mechanical stress, we speculated that the rapid formation of the hematoma exerts pressure on the surrounding cells, causing Ca²⁺-permeable Piezo1 to rapidly respond. Previous research has indicated that Piezo1 activation by traction forces results in Ca²⁺ influx and nuclear localization of the mechanoresponsive transcription factors yes-associated protein and transcriptional coactivator with PDZ-binding motif in neural stem cells (Pathak et al., 2014), resulting in regulation of myelin formation in both Schwann cells (Joseph et al., 2016) and oligodendrocytes (Shimizu et al., 2017). In our study, we found that the myelin sheaths become swollen and broke down in the ipsilateral basal ganglia following ICH, whereas blocking Piezo1 channels using Dooku1 alleviated these effects, strengthening the argument that oligodendrocytes are mechanosensitive and that Piezo1 is a negative regulator of ICH-induced demyelination.

In eukaryotic cells, the ER is essential for Ca²⁺ homeostasis and for protein synthesis, folding, and secretion. ER microenvironment homeostasis is critical for cell activity (Krebs et al., 2015). Previous studies have indicated that an imbalance in Ca²⁺ concentrations in the ER can increase ER stress and activate the unfolded protein response, leading to cell apoptosis (Sovolyova et al., 2014; Spencer and Finnie, 2020). PERK, IRE1, and ATF6, three ER-transmembrane proteins, are involved in the unfolded protein response (Larner et al., 2004). Under ER stress, GRP78 (also known as binding immunoglobulin protein (BiP)) dissociates from three ER stress sensors (IRE1, PERK, and ATF6) (Read and Schröder, 2021), which activates them, enabling them to detect the accumulation of unfolded proteins and initiate the restoration and maintenance of ER homeostasis (Urra et al., 2013). PERK activation causes eIF2α to phosphorylate and activate transcription factor 4 (ATF4), which promotes CHOP expression. CHOP is an important marker of apoptosis that is induced by ER stress and induces the expression of a series of genes that facilitate cell apoptosis, such as GADD34 and ERO1α. Apoptosis also results from immoderate activation of the PERK pathway, which downregulates Bcl-2 expression and upregulates BIM and p53 expression (Hetz and Mollereau, 2014). In response to IRE1α pathway activation, caspase-12 translocates from the ER to the cytosol and cleaves procaspase-9, leading to activation of its effector caspase, caspase-3 (Chiu and Su, 2017). Under ER stress conditions, activated ATF6 is transferred to the Golgi apparatus and cleaved by specific Golgi-resident proteases to form CATF6, which then activates ER stress-related transcription factors to upregulate the expression of BiP and CHOP (Sovolyova et al., 2014). Out results suggest that Piezo1 suppression effectively blocks the PERK-ATF4-CHOP and IRE1 signaling pathways, attenuates ICH-induced oligodendrocyte apoptosis, and alleviates demyelination and neurological impairment. In addition, suppressing Piezo1 channel activity with Dooku1 decreased Bax expression, increased Bcl-2 expression, and increased expression of caspase-12, a characteristic marker of ER stress-related apoptosis. The western blotting results indicated that expression of active caspase-3, the executor of apoptosis, was also decreased by Piezo1 blockade in the ipsilateral basal ganglia after ICH. In addition, TUNEL staining mainly co-localized with immunofluorescence labeling for Olig2, demonstrating that ICH-induced oligodendrocyte apoptosis leads to MBP breakdown and demyelination in an ER pathway-dependent manner. In the future, specific inhibition of these apoptotic pathways could be used as a clinical approach to alleviate oligodendrocyte apoptosis in patients with ICH. Taken together, our findings indicate that Piezo1 suppression markedly decreases the expression of ER stress-related proteins after ICH and significantly decreased oligodendrocyte apoptosis by inhibiting the unfolded protein response.

Our study had several limitations. First, we mainly focused on oligodendrocyte apoptosis and myelin damage after ICH. Axons are also an important component of WM, and the effect of Piezo1 inhibition on neuronal apoptosis and axonal degeneration after ICH should be explored in future research. Second, Piezo1 suppression also has anti-inflammatory effects, as indicated by the changes in inflammatory cytokine levels that we observed in the perihematomal tissues after ICH (Additional Figure 2 ^{(1,020.2KB, tif)} ). Hence, we cannot eliminate the possibility that the anti-inflammatory or other effects also play a role in the neuroprotective effects of Piezo1 suppression. Third, we only evaluated the neuroprotective effects of Piezo1 suppression for 3 days after ICH in this study. The long-term outcomes of Piezo1 suppression after ICH should be assessed in future studies.

In conclusion, we demonstrated for the first time that Piezo1 at least contributes to ICH-induced demyelination and neurological deficits, and that Piezo1 blockade reduces oligodendrocyte apoptosis through the ER pathway after ICH. Therefore, Piezo1 signaling may represent a promising therapeutic target for ameliorating ICH-mediated demyelination.

Additional files:

Additional Figure 1 ^{(2.4MB, tif)} : Entire blot for all western blotting data.

Additional Figure 1

Entire blot for all western blotting data.

ATF: Activating transcription factor; CATF6: cleaved activating transcription factor 6; IF2a: eukaryotic intaitiation factor 2 alpha; IRE1: inositol-requiring enzyme 1; MBP: myelin basic protein; P-eIF2a: phospho-eukaryotic initiation factor 2 alpha; PERK: protein kinase RNA-like endoplasmic reticulum kinase; P-IRE1: phosphor inositol-requiring enzyme 1; P-PERK: phosphor-protein kinase RNA-like endoplasmic reticulum kinase.

Additional Figure 2 ^{(1,020.2KB, tif)} : The effects of Dooku1 on the expression of inflammation factors in the ipsilateral basal ganglia (around the lesion sites) at 24 hours after ICH injury.

Additional Figure 2

The effects of Dooku1 on the expression of inflammation factors in the ipsilateral basal ganglia (around the lesion sites) at 24 hours after ICH injury.

Quantitative analyses of the IL-1β, IL-6 and TNF-α from the results of enzyme-linked immunosorbent assay. Data are expressed as mean ± SD (n = 6). *P < 0.05, vs. sham group; #P < 0.05, vs. ICH + vehicle group (one-way analysis of variance followed by Scheffe's post hoc test). IL: interleukin; TNF-α: tumor necrosis factor α.

Additional file 1 ^{(91.9KB, pdf)} : The number of specific statistics per figure.

Additional file 1

The number of specific statistics per figure