METTL3-dependent N6-methyladenosine modification is involved in berberine-mediated neuroprotection in ischemic stroke by enhancing the stability of NEAT1 in astrocytes

Animals and drug administration

Adult male C57BL/6J mice (weight 20–25 g; 8 weeks old) were gotten from Animal Model Center of Nanjing University. The above mice were placed under strictly controlled ambient temperature (22 ± 2° C) and relative humidity (50~60 %) and 12 h light/dark cycle (five per cage) allowed access to food and water. This study was approved by the Ethics Committee of Nanjing First Hospital (Approval Number of Ethics Committee: DWSY-23063367). All animal experiments were conducted in accordance with the ethical standards of Nanjing First Hospital.

Berberine (purity 98%) was gotten from Sigma (BP1108, USA). Given the present research, we randomly divided all mice into 3 groups, including sham, model (MCAO/R+Vehicle), and model (MCAO/R) + BBR (50 mg/kg). The dose and mode of BBR administration were determined according to our previous studies [25]. The BBR treatment group was pretreated with BBR (dissolved in 0.5% CMC-Na) by intragastric administered once a day for 14 consecutive days before the surgery. An equal dose of normal saline was administered to the other groups.

Middle cerebral artery occlusion/reperfusion (MCAO/R) model mice

Firstly, mice were anesthetized with 2% isoflurane in O₂ (RWD Life Science, Shenzhen, China). A neck incision was performed, exposing the right common carotid artery (CCA), external carotid artery (ECA), and internal carotid artery (ICA). After that, a silicon-coated monofilament (Cinontech, Beijing, China) was inserted into the CCA and advanced into the ICA, then, occluded the artery for 1h followed by reperfusion. The same surgical procedure was performed on sham group of mice, while without a monofilament put into the artery.

Neurologic function scores

At 24 h after MCAO/R, neurological deficits of mice were assessed following the method described by Longa et al. [26]. The scoring criteria are as follows: 0, no deficit, normal walking; 1, mild deficit, fail to stretch forepaw; 2, moderate deficit, circling to the contralateral side; 3, severe deficit, falling to the contralateral side; 4, no spontaneous moving.

2,3,5-triphenyltetrazolium chloride (TTC) staining and infarct volume measurement

Triphenyl tetrazolium chloride (TTC) staining was performed at 24 h after surgery. Firstly, the brains were cut into 6 sections. Then, sections were stained using 2% TTC solution (Sigma-Aldrich, St. Louis, MO, USA) at 37° C for 0.5h, and fixed with 4% paraformaldehyde (Biosharp, Beijing, China). ImageJ software (National Institutes of Health, Bethesda, MD, USA) was used to calculate the percentages of normal tissue (red) and infarcted volume (white). The infarcted areas of each section were summed and presented as a percentage of the volume of un-infarcted areas: corrected percentage of infarct volume = [(contralateral hemisphere volume − ipsilateral non-infarct volume)/contralateral hemisphere volume] × 100%.

Cell culture

Choosing the cerebral cortex of mice (1-3 days old) to extract primary astrocytes, the cortex was digested by 0.125% trypsin-EDTA (Gibco, Grand Island, NY, USA), then centrifuged and incubated in DMEM/F12 complete medium (Gibco, Thermo Fisher Scientific, Waltham, MA, USA). After 9-11 days, the cells were shaken at 260 rpm at 37° C for 16 h to purify. Finally, purified astrocytes were obtained.

Choosing the cortex of fetal mice (16–18 days old) to isolate primary neurons, the culture flasks were pretreated with poly-d-lysine (PDL) (Sigma-Aldrich, St. Louis, MO, USA). The fragment was digested with 0.125% trypsin and grown with DMEM (Gibco, Grand Island, NY, USA) for 5 h, then replaced with neurobasal medium containing 2% B27 (Gibco, Grand Island, NY, USA) and 0.5 mmol/L glutamine (Sigma-Aldrich, St. Louis, MO, USA).

Choosing Adult Brain Dissociation Kit and ACSA-2 MicroBeads (Miltenyi Biotec, Bergisch Gladbach, Germany) for magnetic isolation of astrocytes from adult mice. The former was used to digest the cortex. Then, purified astrocytes were obtained. Astrocytes were incubated for 15 min at 4° C with ACSA-2 MicroBeads and separated from single-cell suspension in a magnetic field using MS columns, MACS MultiStand and QuadroMACS (Miltenyi Biotec, Bergisch Gladbach, Germany).

Primary astrocytes were co-cultured with primary neurons

The transwell co-culture system was established as described [27, 28]. After transfection for 48h, the medium was discarded and the primary astrocytes were washed by PBS. The primary neurons were then placed at the bottom of the transwell plate (Corning Company, Corning, NY, USA). After washing with PBS, primary astrocytes were put in the upper part of the transwell plate and separated from bottom by a semi-permeable membrane. The cells were incubated in two separate chambers.

Oxygen-glucose deprivation/reoxygenation (OGD/R) treatment

Cells were incubated with or without BBR (10 μM) for 24 h, then exposed to OGD/R treatment. So as to mimic the environment of cerebral ischemia/reperfusion in vitro, cells exposed glucose and oxygen deprivation followed reoxygenation. First, cells were cultured with DMEM (Gibco, Grand Island, NY, USA) without glucose and FBS (Gibco, Grand Island, NY, USA) in a 95% N₂ and 5% CO₂ chamber for 4 h. Then the glucose-free DMEM medium was replaced with neurobasal medium complete medium or DMEM/F12 complete medium, and cultured in a 95% air and 5% CO₂ incubator for 24 h. The final concentration of BBR in the OGD/R experiment was 10 μM.

Cell transfection and lentivirus infection

Briefly, miR-377-3p mimic, miR-377-3p inhibitor, and respective negative control (NC mimic, NC inhibitor) were purchased from Shanghai GenePharma Company (Shanghai, China). According to the manufacturer’s protocol, transfected them to cells with Lipofectamine 3000 (Invitrogen, Carlsbad, CA, USA). Lentiviruses loaded with shRNA were obtained from Shanghai GeneChem Company (Shanghai, China), including overexpressed METTL3 (Lenti-METTL3), sh-METTL3, overexpressed NEAT1 (Lenti-NEAT1), sh-NEAT1 and their corresponding lentiviral scramble control shRNA or their corresponding overexpression negative control vector (Lenti-NC). Lentivirus infection was performed according to the manufacturer’s instructions. The sequences above are shown in Supplementary Table 1. After transfection or lentiviral infection, cells were treated with or without 10 μM of BBR and then subjected to subsequent OGD/R treatment to simulate the brain ischemia-reperfusion environment in vitro.

Cell counting kit-8 (CCK-8)

Cell Counting Kit-8 (Dojindo, Kumamoto, Japan) was used to detect cell viability. Firstly, cells were treated with 10 μL CCK-8. Finally, after culture for 2 h at 37° C, the absorbance value at 450 nm was measured using a microplate reader (Thermo Fisher Scientific, Waltham, MA, USA). The survival ratio of cells per group was normalized to the control group, in which the survival ratio was regarded as 100%.

Lactate dehydrogenase (LDH) assay

Cellular injury or death was detected by the release of lactate dehydrogenase (LDH) in the medium supernatant. First, collecting cells treated with OGD/R and transfected, put them suspended in 96-well plate, and then placed at 37° C for 0.5 h. Choosing a LDH assay kit (Sigma-Aldrich, St. Louis, MO, USA) to carry out the experiment. The absorbance of the supernatant at 490 nm was then recorded using a microplate reader. The LDH activity was calculated as (absorbance of sample hole-absorbance of the control hole) / (absorbance of the standard hole - absorbance of the standard blank hole).

HE staining

24 h after MCAO/R, mice were anesthetized with 2% isoflurane in O₂ (RWD Life Science, Shenzhen, China). After that normal saline (200 mL) and 4% paraformaldehyde (Biosharp, Beijing, China) (80 mL) were perfused into the heart. Brain tissue samples were fixed with 4% paraformaldehyde, dehydrated and embedded in paraffin. 4 μm slices were cut with a paraffin microtome and attached to the slide. Dewaxing by xylene and gradient dehydration by ethanol were adopted. At room temperature, hematoxylin staining solution was used for 5 min, followed by 1% ethanol hydrochloride for 30s differentiation. The ammonia water was added for 1 minute and then turn blue. They were rinsed with distilled water. Eosin staining solution was added, room temperature 2 min, distilled water washed, ethanol gradient decolorization. The xylene was allowed to permeate. Finally, the slide is sealed with a neutral adhesive. Changes were observed and photographed.

Nissl staining

Similarly, the mice were anesthetized by 2% isoflurane in O₂ (RWD Life Science, Shenzhen, China), brain tissue were fixed with 4% paraformaldehyde, dehydrated and embedded in paraffin. 4 μm slices were cut with a paraffin microtome and attached to the slide. Then carried out the Nissl staining assay according to the protocol of the Nissl staining kit (Solarbio, Nanjing, China). The results were viewed and photographed.

Immunofluorescence

The isolated astrocytes or neurons were characterized by staining marker GFAP or Neun. Firstly, cells were put in 4% paraformaldehyde (Biosharp, Beijing, China) for 20 min. After that, put them in PBS containing 0.1% Triton X100 (Beyotime, Shanghai, China) for 20 min. Then incubated them with 5 % BSA (Biofroxx, Shanghai, China) for 1 h. After above, primary astrocytes were incubated with rabbit monoclonal antibody anti-GFAP (1:200; #80788S; CST, USA) and primary neurons were incubated with a rabbit monoclonal antibody anti-NeuN (1:200; #24307; CST) overnight at 4° C and then incubated the next day with goat anti-rabbit IgG H&L (Alexa Fluor® 488) (1:200; ab150077; Abcam, Cambridge, MA, USA) or goat anti-rabbit IgG H&L (Alexa Fluor® 594) (1:200; ab150080; Abcam) and incubated with DAPI (1:300, Cat. # C1002, Beyotime) to stain nuclei. Finally, the images were observed and photographed. The number of GFAP-positive cells or Neun-positive cells was calculated from randomly selected microscopic fields. Astrocyte or neuronal purity was presented as percentage of the number of the percentage of GFAP-positive or Neun-positive cells to the total cells of view, both around 90% or more.

mRNA stability assay

The cells were planted in 6-well plates overnight. Cells were incubated with 5 μg/ml ActD (CST, Danvers, MA, USA) for 0, 1, 2, 4 h or longer. Then collected cells to extract total RNA, and target RNA was quantized by RT-qPCR for further analysis.

Bioinformatic analysis

The possible miRNAs-related targets of Nampt were analyzed by online bioinformatics database miRDB, TargetScan and miRWalk. And through bioinformatics software RNAhybrid (https://bibiserv.cebitec.unibielefeld.de/rnahybrid/) to assess the combine about miR-377-3p with LncNEAT1. The bioinformatics software m6Avar (http://www.cuilab.cn/sramp/) was used to analyze the potential m6A sites on LncNEAT1.

Dual-luciferase reporter gene assay

First, the luciferase reporter plasmids Nampt-WT and Nampt-Mut was constructed, then cells were co-transfected with above plasmids and NC mimic or miR-377-3p mimic respectively by using Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA). After co-incubation for 48h, cells were collected for lysis and luciferase activity was measured by the dual luciferase reporting and detection system (Promega, WI, USA). Similarly, the protocol also used to explore the combined effects of NEAT1 and miR-377-3p.

Methylated RNA immunoprecipitation (MeRIP) assay

Firstly, Trizol method was used to extract total RNA. The RNA cracking reagent (AM8740, Invitrogen) was used to cut RNA into fragments. Anti-m6a antibodies (ab208577, Abcam) were incubated with RNA overnight at 4° C. Protein A/G Magnetic Beads (88803, Thermo Fisher Scientific, USA) was mixed with antibody treated RNA. A protease K buffer digested the m6A antibody. Finally, the methylated RNA was purified for RT-qPCR.

Western blotting

Firstly, RIPA solution (Beyotime, Shanghai, China) was used to extract total proteins from mouse brain tissue or cells. Total protein concentration was measured and quantified using a BCA kit (Thermo Fisher Scientific, Grand Island, NY, USA). Protein samples mixed with 5×loading buffer were denatured in boiling water, separated by gel electrophoresis and transferred to PVDF membrane (Merck Group, Darmstadt, Germany). Closed with 5% skim milk for 1h and incubated with the following primary antibodies at 4° C overnight: rabbit monoclonal anti-Nampt (1:1000; 11776-1-AP; Proteintech, China), β-actin (1:1000; ab241153; Abcam), rabbit monoclonal anti-METTL3 (1:1000; 15073-1-AP; Proteintech). PVDF membrane is then paired with corresponding HRP labeled anti-Rabbit IgG (1:4000; #7074S; CST) was incubated at room temperature for 2h. Strip visualization was done using the ECL kit (Thermo Fisher Scientific, USA). The results were quantified using software ImageJ.

Quantitative real-time polymerase chain reaction (RT-qPCR)

Trizol reagent (Thermo Fisher Scientific, USA) extracts total RNA from tissues or cells. RNA samples were reversely transcribed into cDNA using kits (ABM, Richmond, British Columbia, Canada). According to the standard SYBR-Green method, the ABI7500 type sequence detection system (7500, ABI, USA) was used for the detection. β-actin being an internal control of mRNA, and U6 was an internal reference of miRNA. The primers are shown in Supplementary Table 2.

Statistical analysis

These experiments were performed at least three times. All values are expressed as mean ± standard deviation. Prism 8.0 (GraphPad Software, Inc., San Diego, CA, USA) should be used in all statistical analysis. All data were normally distributed. The difference between the two groups was checked by the T-test, and the comparison between more than two groups was checked by the analysis of variance. P<0.05 was regarded as statistically significant.

Data availability statement

All data generated or analyzed during this study are available from the corresponding author upon reasonable request.

Berberine alleviates MCAO/R brain injury and up-regulates the expression of Nampt in adult mouse astrocytes

Firstly, the mice were put to MCAO/R to simulate cerebral I/R injury in vivo. Compared with the sham group, the nerve function defect and infarct volume of MCAO/R mice were strikingly increased, and the above results were reversed after administration of BBR (Figure 1A–1C). Meanwhile, HE staining and Nissl staining were used to observe the degree of brain histological damage in mice. The results indicated that ischemic stroke resulted in shrinkage of nucleus of neurons, widened pericellular spaces, as well as vacuolization in neuropil and decrease of Nissl bodies, while berberine administration relatively reduced brain damage in mice (Figure 1D–1G). Then we isolated astrocytes from 8-week-old adult mice. Western blot analysis demonstrated that compared with the sham group, the expression of Nampt in astrocytes of MCAO/R mice brain tissue was increase, while BBR administration was further increased the expression of Nampt (Figure 1H, 1I). To sum up, these results suggest that berberine may reduce the brain injury of mice after MCAO/R and act as a protective role against I/R injury by up-regulating the expression of Nampt in mouse astrocytes.

Figure 1. Berberine alleviates MCAO/R brain injury and up-regulates the expression of Nampt in adult mouse astrocytes. (A) Neurologic function scores for assessing the degree of neurological impairment after 24 hours in sham, MCAO/R+Vehicle or BBR+ MCAO/R group of mice (n=24). (B) Representative images of TTC-stained brain sections from mice (n=6). (C) Quantification of infarct volume at 24 h after MCAO/R (n=6). (D) Representative images of HE staining of ischemic brain tissue from mice (n=6, Scale bar: 50 μm). (E) Representative images of Nissl staining of ischemic brain tissue from mice (n=6, Scale bar: 50 μm). (F) Quantification of damaged neurons. (G) Quantification of Nissl+ neurons. (H) Representative Western blot images depicting Nampt in adult mouse astrocytes (n=6). (I) A bar presenting the quantification of Nampt (n=6). The relative expression levels were quantified by normalizing to β-actin. Data are represented as mean ± SD, (*P < 0.05; **P < 0.01; ***P < 0.001).

miR-377-3p targets and inhibits the Nampt expression in OGD/R-treated primary astrocytes, thus promoting neuronal injury

Above three online bioinformatics databases were used to perform the potential miRNAs that can bind to 3’-UTR of Nampt were screened and miR-377-3p was initially determined (Figure 2A). Then, we extracted mouse primary astrocytes (Figure 2B). In order to further confirm the direct binding of Nampt and miR-377-3p, we conducted dual-luciferase reporter gene assay. As expected, miR-377-3p dramatically reduced the activity of Nampt-WT luciferase, but did not reduce the activity of Nampt-mut (Figure 2C). In order to further evaluate the interaction between Nampt and miR-377-3p, we silenced or over-expressed the level of miR-377-3p in primary astrocytes of mice and then treated by OGD/R (Figure 2D). The results showed that after OGD/R, compared with the corresponding control group, overexpression of miR-377-3p prominently reduced the level of Nampt, while knocking down miR-377-3p distinctly increased the level of Nampt (Figure 2E, 2F). To determine the influence of miR-377-3p in neuroprotective effects associated with Nampt, we extracted the primary neurons of mice (Figure 2G). After primary astrocytes were transfected with miR-377-3p mimic or miR-377-3p inhibitor, they were co-cultured with primary neurons, and then the system treated by OGD/R. CCK8 assay indicated that inhibition of miR-377-3p in primary astrocytes can promote the activity of its co-cultured neurons after OGD/R, while miR-377-3p mimic decreased the neuronal activity (Figure 2H). At the same time, the LDH experiment results also revealed that compared with the corresponding control group, the primary astrocyte transfected with miR-377-3p inhibitor attenuated the death of neurons co-cultured with it after OGD/R, while the transfection of miR-377-3p mimic increased the death of neurons (Figure 2I). Generally, our work confirm that the neuronal damage mediated by OGD/R-treated primary astrocytes is related to its Nampt/miR-377-3p signal transduction.

Figure 2. miR-377-3p inhibits the expression of Nampt in astrocytes after OGD/R, thus promoting neuronal injury. (A) Possible miRNA binding site on Nampt mRNA (predicted candidate miR-377-3p target gene by bioinformatics databases). (B) GFAP immunofluorescence validation of astrocytes (immunofluorescence images of primary astrocytes. Scale bar: 20 μm). (C) Relative luciferase activity of Nampt wild-type and 3ʹ-UTR mutant structures transfected with miR-377-3p mimics and NC mimic. (D) RT-qPCR detection of miR-377-3p levels in primary astrocytes after OGD/R following miR-377-3p mimic or miR-377-3p inhibitor treatment. (E) Western blot analysis of Nampt protein expression in primary astrocytes after OGD/R following miR-377-3p mimic or miR-377-3p inhibitor treatment. (F) A bar presenting the quantification of Nampt in primary astrocytes. (G) NeuN immunofluorescence validation of neurons (immunofluorescence images of primary neurons. (Scale bar: 20 μm.). (H) The co-cultured neuron viability was determined by CCK-8 assay after miR-377-3p mimic or miR-377-3p inhibitor transfection in astrocytes and OGD/R treatment. (I) LDH assay to detect the effect on co-cultured neurons after miR-377-3p mimic or miR-377-3p inhibitor treatment in primary astrocytes and OGD/R treatment. The relative expression levels were quantified by normalizing to β-actin. Data are represented as mean ± SD, (n = 3; *P < 0.05; **P < 0.01; ***P < 0.001).

Berberine regulates Nampt expression through downregulation of miR-377-3p in OGD/R-treated primary astrocytes, thereby affecting neuronal activity

Subsequently, we wanted to further investigate whether berberine plays a neuroprotective role by regulating Nampt-related pathways in primary astrocytes. RT-qPCR results nominated that miR-377-3p was down-regulated in primary astrocytes after OGD/R compared to control. Compared with the corresponding control group, miR-377-3p in OGD/R-induced primary astrocytes was further down-regulated after berberine treatment. Meanwhile, overexpression of miR-377-3p in OGD/R-treated primary astrocytes reversed the decrease of miR-377-3p induced by berberine (Figure 3A). Western blot analysis revealed that Nampt expression was up-regulated in post-OGD/R primary astrocytes. Compared with the corresponding control group, after berberine treatment, Nampt expression was further up-regulated in OGD/R-induced primary astrocytes, and overexpression of miR-377-3p in OGD/R-treated astrocytes reversed the up-regulation of Nampt expression induced by berberine (Figure 3B, 2C). Significantly, compared with the corresponding control group, berberine treatment can attenuate the neuronal damage induced by primary astrocytes after OGD/R. Nevertheless, transfection of miR-377-3p mimic reversed the neuroprotective effect induced by BBR treatment in primary astrocytes after OGD/R, which can be demonstrated by the decreased neuronal activity and increased LDH release (Figure 3D, 3E). In conclusion, berberine plays a neuronal protective role by down-regulating the expression of miR-377-3p through Nampt in primary astrocytes.

Figure 3. Berberine regulates Nampt expression through downregulation of miR-377-3p in OGD/R-treated primary astrocytes, thereby affecting neuronal activity. (A) Using RT-qPCR to analysis the miR-377-3p level in cells transfected with miR-377-3p mimic after berberine treatment. (B) Using Western blot to analyze Nampt expression in astrocytes after berberine treatment and transfection with miR-377-3p mimic. (C) A bar presenting the quantification of Nampt in primary astrocytes. (D) Using the CCK-8 assay to evaluate the effect of co-culture with berberine and primary astrocyte transfected with miR-377-3p mimic on neurons after OGD/R. (E) Using LDH assay to study the effect of co-culture with berberine and primary astrocyte transfected with miR-377-3p mimic on neurons after OGD/R. The relative expression levels were quantified by normalizing to β-actin. Data are represented as mean ± SD, (n = 3; *P < 0.05; **P < 0.01; ***P < 0.001).

NEAT1 binds to miR-377-3p as a ceRNA in primary astrocytes and plays a neuronal protective role

It is reported that LncNEAT1 can alleviate cerebral ischemia/reperfusion injury [29]. Firstly, the results assessed by the bioinformatics software RNAhybrid (https://bibiserv.cebitec.unibielefeld.de/rnahybrid/) indicated that miR-377-3p has a strong ability to bind to NEAT1 (Figure 4A). Meanwhile, the dual-luciferase reporter gene assay results confirmed that miR-377-3p-mimic markedly down-regulated the luciferase activity of NEAT1-WT, but did not down-regulate the luciferase activity of NEAT1-Mut (Figure 4B). In order to further evaluate the interaction between NEAT1 and miR-377-3p, we knocked-down or over-expressed the level of NEAT1 in primary astrocytes of mice and then treated by OGD/R (Figure 4C). Next, our results revealed that overexpression of NEAT1 significantly decreased the level of miR-377-3p in primary astrocytes, while knocking down NEAT1 prominently increased the level of miR-377-3p (Figure 4D). The primary astrocytes were knocked-down or overexpressed NEAT1 and co-cultured with the primary neurons, then the system treated by OGD/R. CCK8 assay demonstrated that compared with the corresponding control group, overexpression of NEAT1 in primary astrocytes promoted the activity of its co-cultured neurons exposed to OGD/R treatment, while knocking down NEAT1 weakened the neuronal activity (Figure 4E). At the same time, the results of LDH assay also indicated that compared with the corresponding control group, after OGD/R, the NEAT1 overexpression in primary astrocytes reduced the death of co-cultured neurons, while knocking down NEAT1 increased the death of neurons (Figure 4F). In general, the above results showed that NEAT1 binds to miR-377-3p as a ceRNA in primary astrocytes and plays a neuronal protective role.

Figure 4. NEAT1 binds to miR-377-3p in OGD/R astrocytes and plays a neuronal protective role as a ceRNA. (A) Prediction of NEAT1 binding sites on miR-377-3p using bioinformatics software (RNAhybrid). (Schematic diagram of the sequence complementation relationship of miR-377-3p in NEAT1.) (B) Dual luciferase reporter gene analysis to confirm the binding relationship between NEAT1 and miR-377-3p. (C) RT-qPCR to analyze the NEAT1 level in primary astrocytes transfected with sh-NEAT1 or Lenti- NEAT1. (D) RT-qPCR to analyze the miR-377-3p level in primary astrocytes transfected with sh-NEAT1 or Lenti- NEAT1. (E) CCK-8 assay to assess the co-cultured neuron viability after NEAT1 knockdown or overexpression in primary astrocytes and OGD/R treatment. (F) LDH assay to study the effect of the co-cultured neuron viability after NEAT1 knockdown or overexpression in primary astrocytes and OGD/R treatment. Data are represented as mean ± SD, (n = 3; *P < 0.05; **P < 0.01; ***P < 0.001).

Berberine exerts neuroprotective effects through modulation of the NEAT1/ miR-377-3p /Nampt axis in primary astrocytes

Next, we wanted to further explore whether berberine exerts neuroprotective effects by regulating NEAT1/miR-377-3p/Nampt-related pathway in primary astrocytes. RT-qPCR results confirmed that NEAT1 was upregulated in primary astrocytes after OGD/R compared to control, and NEAT1 was further upregulated in primary astrocytes after OGD/R following berberine treatment compared to the corresponding control group, while sh-NEAT1 transfection of primary astrocytes reversed the increase of NEAT1 expression induced by berberine treatment (Figure 5A). Meanwhile, the results nominated that miR-377-3p was downregulated in primary astrocytes after OGD/R following berberine treatment compared to the corresponding control group, while sh-NEAT1 transfection in primary astrocytes reversed the decrease in miR-377-3p induced by berberine treatment (Figure 5B). Western blot analysis revealed that Nampt expression was upregulated in primary astrocytes after OGD/R following berberine treatment compared to the corresponding control group, and sh-NEAT1 transfection of primary astrocytes reversed the increase in Nampt induced by berberine treatment (Figure 5C, 5D). Notably, transfection of sh-NEAT1 in primary astrocytes reversed the neuroprotective effect induced by berberine treatment compared to the corresponding controls, as evidenced by the decreased neuronal activity and increased LDH release (Figure 5E, 5F). Taken together, it can be known that LncRNA NEAT1 can be regarded as a competitive endogenous RNA (ceRNA) for miR-377-3p and block miR-377-3p-dependent target gene repression (i.e. Nampt). Concurrently berberine can upregulate NEAT1 and further regulate Nampt-related signaling in post-OGD/R primary astrocytes, thereby affecting neuronal activity.

Figure 5. Berberine exerts neuroprotective effects through modulation of the NEAT1/ miR-377-3p /Nampt axis in astrocytes. (A) RT-qPCR to analyze the NEAT1 level in primary astrocytes transfected with sh-NEAT1 after berberine treatment. (B) RT-qPCR to analyze the miR-377-3p level in primary astrocytes transfected with sh-NEAT1 after berberine treatment. (C) Western blot analysis of Nampt expression in primary astrocytes transfected with sh-NEAT1 after berberine treatment. (D) A bar presenting the quantification of Nampt in primary astrocytes. (E) CCK-8 assay to assess the effect on co-cultured neurons by treating with NEAT1 knockdown and berberine-treated primary astrocytes and OGD/R treatment. (F) LDH assay to study the effect on co-cultured neurons by treating with NEAT1 knockdown and berberine-treated primary astrocytes and OGD/R treatment. The relative expression levels were quantified by normalizing to β-actin. Data are represented as mean ± SD, (n = 3; *P < 0.05; **P < 0.01; ***P < 0.001).

METTL3 enhances NEAT1 stability in primary astrocytes by modulating m6A modification of NEAT1

To probe the possible mechanisms about the regulation of neuronal protective effects by NEAT1, we examined the N6-methyladenosine (m6A) methylation of NEAT1. We used the bioinformatics software m6Avar to analyze potential m6A sites on LncNEAT1 and found multiple positions on LncNEAT1 with high confidence (Figure 6A). In addition, MeRIP-qPCR results confirmed that following OGD/R stimulation, LncNEAT1 m6A methylation levels were significantly increased in primary astrocytes (Figure 6B).

Figure 6. METTL3 enhances NEAT1 stability in OGD/R astrocytes by modulating m6A modification of NEAT1. (A) Prediction of potential m6A targets in Lnc NEAT1 based on SRAMP online website. (B) MeRIP-PCR to detect the m6A modifications in Lnc NEAT1. (C) Application of MeRIP-PCR to detect the correlation between NEAT1 and METTL3. (D) Western blot to detect expression of METTL3 after overexpression or silencing in primary astrocytes. (E) A bar presenting the quantification of METTL3 in primary astrocytes. (F) Expression of NEAT1 after METTL3 overexpression and silencing in primary astrocytes of mice after OGD/R. (G) Stability of NEAT1 in primary astrocytes with METTL3 overexpression after OGD/R treatment. Data are represented as mean ± SD, (n = 3; *P < 0.05; **P < 0.01; ***P < 0.001).

Published studies have revealed that methylation transferase METTL3 can increase NEAT1 expression by promoting m6A modification of NEAT1 [30]. We assessed the possibility that methyltransferase METTL3 is involved in regulating NEAT1 m6A modification in ischemic stroke by MeRIP-qPCR. The result indicated that knockdown of methyltransferase METTL3 apparently decreased m6A modification on NEAT1 (Figure 6C). Next, to further investigate the effect of METTL3 on NEAT1 expression levels, we knocked down or overexpressed METTL3 in primary astrocytes of mice and then treated with OGD/R (Figure 6D, 6E). Furthermore, RT-qPCR results revealed that the level of NEAT1 was significantly up-regulated in primary astrocytes after OGD/R treatment with overexpressing METTL3 and down-regulated in METTL3 knockdown cells compared to the corresponding controls (Figure 6F). To further certify the rationality of METTL3-mediated NEAT1 regulation, we investigated the effect of METTL3 overexpression on NEAT1 stability. Overexpressed of METTL3 in primary astrocytes significantly increased the stability of NEAT1 early after OGD/R (4h) upon ActD treatment (Figure 6G). The above findings reveal that METTL3 enhances NEAT1 stability through m6A modification in OGD/R-treated primary astrocytes.

Berberine exerts neuroprotective effects via METTL3 regulating the NEAT1/miR-377-3p/Nampt axis in post-OGD/R primary astrocytes

In view of the above studies on the neuroprotective role of berberine in regulating NEAT1-related pathways in post-OGD/R primary astrocytes, we further investigated the effect of berberine on m6A modification in an in vitro model of ischemic stroke. Western blot results showed that METTL3 expression was upregulated after OGD/R compared with control group, besides, METTL3 expression was further upregulated in post-OGD/R primary astrocytes after berberine treatment compared with OGD/R group (Figure 7A, 7B). Meanwhile, knockdown of METTL3 in primary astrocytes after OGD/R reversed the increase of NEAT1 induced by berberine treatment (Figure 7C), but conversely made miR-377-3p expression up-regulated (Figure 7D). In contrast, compared with the corresponding control group, knockdown of METTL3 in primary astrocytes after OGD/R reversed the increase of Nampt induced by berberine treatment (Figure 7E, 7F). Notably, knockdown of METTL3 also reversed the neuronal protective effects induced by berberine treatment on OGD/R primary astrocytes compared to the corresponding controls, as evidenced by the relative reduction in neuronal viability and increased LDH release (Figure 7G, 7H). Taken together, we can know that berberine exerts neuroprotective effects via METTL3 regulation of the NEAT1/miR-377-3p/Nampt axis in post-OGD/R primary astrocytes.

Figure 7. Berberine exerts neuroprotective effects via METTL3 regulating the NEAT1/miR-377-3p/Nampt axis in post-OGD/R primary astrocytes. (A) Western blot to detect expression of METTL3 in primary astrocytes after OGD/R and berberine treatment. (B) A bar presenting the quantification of METTL3 in primary astrocytes. (C) RT-qPCR to analysis the ex-pression of NEAT1 in primary astrocytes after OGD/R following berberine treatment and transfection with sh-METTL3. (D) RT-qPCR to analyze the expression of miR-377-3p in astrocytes after OGD/R following berberine treatment and transfection with sh-METTL3. (E) Western blotting to analyze the expression of Nampt in astrocytes after OGD/R following berberine treatment and transfection with sh-METTL3. (F) A bar presenting the quantification of Nampt in primary astrocytes. (G) Using CCK-8 assay to assess the effect on co-cultured neurons by treating with sh-METTL3 and berberine-treated primary astrocytes and OGD/R treatment. (H) Using LDH assay to evaluate the effect on co-cultured neurons by treating with sh-METTL3 and berberine-treated primary astrocytes and OGD/R treatment. The relative expression levels were quantified by normalizing to β-actin. Data are represented as mean ± SD, (n = 3; *P < 0.05; **P < 0.01; ***P < 0.001).

Berberine regulates METTL3-mediated m6A modification of NEAT1 to alleviate ischemic stroke in adult mouse astrocytes

Finally, in view of the above experiments, we demonstrated the neuronal protective effect of berberine in an in vitro model of cerebral I/R and the specific correlated mechanism, so we wanted to further validate the neuronal protective effect mechanism of berberine in mice in vivo. Then we isolated astrocytes from 8-week-old adult mice. Western blot analysis confirmed that the expression of METTL3 was upregulated in astrocytes of brain tissue from MCAO/R mice compared with the sham group, and compared with the model group, BBR administration further increased METTL3 expression (Figure 8A, 8B). Meanwhile, RT-qPCR results further showed that the level of NEAT1 was upregulated in astrocytes of brain tissue from MCAO/R mice compared with the sham group, and further increased by BBR treatment (Figure 8C). In contrast, miR-377-3p levels were downregulated in astrocytes of brain tissue from MCAO/R mice compared to the sham group, and miR-377-3p expression further decrease by BBR administration compared to the model group (Figure 8D). These findings are consistent with the above in vitro results, suggesting that berberine exerts neuroprotective effects by regulating NEAT1/ miR-377-3p/Nampt axis in mouse astrocytes through METTL3, thereby alleviates ischemic stroke.

Figure 8. Berberine regulates METTL3-mediated m6A modification of NEAT1 to alleviate ischemic stroke in adult mouse astrocytes. (A) Western blot to verify METTL3 expression in adult mouse astrocytes after berberine administration following MCAO/R. (B) A bar presenting the quantification of METTL3. (C) RT-qPCR for NEAT1 expression in adult mouse astrocytes after berberine administration following MCAO/R. (D) RT-qPCR to verify the expression of miR-377-3p in adult mouse astrocytes after berberine administration following MCAO/R. The relative expression levels were quantified by normalizing to β-actin. Data are represented as mean ± SD, (n = 6; **P < 0.01; ***P < 0.001).