Puerarin attenuates intracerebral hemorrhage‐induced early brain injury possibly by PI3K/Akt signal activation‐mediated suppression of NF‐κB pathway

Abstract Intracerebral hemorrhage (ICH) can induce intensively oxidative stress, neuroinflammation, and brain cell apoptosis. However, currently, there is no highly effective treatment available. Puerarin (PUE) possesses excellent neuroprotective effects by suppressing the NF‐κB pathway and activating the PI3K/Akt signal, but its role and related mechanisms in ICH‐induced early brain injury (EBI) remain unclear. In this study, we intended to observe the effects of PUE and molecular mechanisms on ICH‐induced EBI. ICH was induced in rats by collagenase IV injection. PUE was intraperitoneally administrated alone or with simultaneously intracerebroventricular injection of LY294002 (a specific inhibitor of the PI3K/Akt signal). Neurological deficiency, histological impairment, brain edema, hematoma volume, blood–brain barrier destruction, and brain cell apoptosis were evaluated. Western blot, immunohistochemistry staining, reactive oxygen species (ROS) measurement, and enzyme‐linked immunosorbent assay were performed. PUE administration at 50 mg/kg and 100 mg/kg could significantly reduce ICH‐induced neurological deficits and EBI. Moreover, PUE could notably restrain ICH‐induced upregulation of the NF‐κB pathway, pro‐inflammatory cytokines, ROS level, and apoptotic pathway and activate the PI3K/Akt signal. However, LY294002 delivery could efficaciously weaken these neuroprotective effects of PUE. Overall, PUE could attenuate ICH‐induced behavioral defects and EBI possibly by PI3K/Akt signal stimulation‐mediated inhibition of the NF‐κB pathway, and this made PUE a potential candidate as a promising therapeutic option for ICH‐induced EBI.


| Animals
Male adult rats (Sprague-Dawley, 280-320 g) were purchased from the Animal Experiment Center of Southern Medical University

| Experimental design and group
In our research, three experiments were designed and orderly performed. The specific experimental design and group were detailed in Supplementary materials, respectively (Fig. S1, Fig. S2, Fig. S3).

| Rat ICH model
ICH was induced in rats as mentioned in previous reports. 44,46,47 Simply, after anesthetized with pentobarbital sodium (45 mg/kg) (Cat. No.: P3761, Sigma-Aldrich, St. Louis, MO, USA), a middle scalp incision was made in rats to expose the bregma. Next, a burr hole (1 mm in diameter) was drilled on the right skull, and a microsyringe (5 μl) was directly inserted into the right striatum (bregma: lateral 3.5 mm, anterior 0.1 mm, and ventral 6.0 mm). Collagenase IV (1 μl, 0.2 U/μl in 0.9% normal saline) (Cat. No.: C1889, Sigma-Aldrich, St. Louis, MO, USA) was injected after 10 min. After in situ for another 10 min, the syringe was slowly moved out. Sham group rats were dealt with an identical method, except that only 1 μl of 0.9% normal saline was injected. The rats were then placed in separate cages and were provided with standard food and water.

| Behavioral deficiency
The modified neurological severity score (mNSS) scale was applied to estimate the neurobehavioral deficiency at 24 h and 72 h following ICH, which was executed by two experienced researchers who were blinded to animal groups. 44,45,50,51 The scale is composed of sensory, motor, reflex, and balance tests, and a higher score means more severe neurological injury. 44,45,50,51

| Paraffin sections
Paraffin sections of rat brains were made, as previously described. [44][45][46]52 In brief, after transcardial perfusion with PBS followed by 4% paraformaldehyde, the rat brains were taken out and performed postfixation in the same fixation solution (4°C, 24 h). After being dehydrated and vitrified, the brain samples were embedded into paraffin. Then, after being dewaxed and rehydrated, the brain sections (4 μm thickness) were used to conduct hematoxylin and eosin (H&E), immunohistochemistry (IHC), and terminal deoxynucleotidyl transferase-mediated biotinylated dUTP nick-end labeling (TUNEL) staining.

| H&E staining
H&E staining was performed as our previous method. 44,46 In brief, the prepared brain sections were immersed into eosin for 10 s and then into hematoxylin for 5 min. After dealing with graded ethanol and xylene, the brain sections were mounted and imaged with a microscope (DM2500, Leica, Germany).

| Hematoma volume
Hematoma volume was assessed as previously reported with some modifications. 53,54 In brief, serially coronal sections (2 mm thickness) of rat brains were prepared and imaged with a digital camera. Then, the acquired images were utilized to calculate the hematoma volume with the following formula: V = T 1 *S 1 + T 2 *S 2 + … + T n *S n [V: hematoma volume (mm 3 ), T: slice thickness, S: hematoma area, n: serial number of brain slices].

| Brain water content
Brain water content (BWC) was measured by the wet/dry weight method, as previously described. 44,55,56 In brief, at 24 h or 72 h following ICH, the rat brains were harvested and separated into five parts (ipsilateral and contralateral cerebral cortices, ipsilateral and contralateral basal ganglia tissues, and cerebellum). After wet weight was obtained, the brain samples were dried to get dry weight (100°C, 24 h). BWC was calculated with the following formula: [(M 1 −M 2 )]/M 1 *100% (M 1 : wet weight, M 2 : dry weight).

| Blood-brain barrier
Evans blue (EB) dye (Wako Pure Chemical Industries, Japan) was applied to assess the disruption of the blood-brain barrier (BBB), as reported previously. 44,52,57 In brief, after intravenous injection of 2% EB for 2 h, the rats were transcardially perfused and then right brain hemispheres were isolated. Next, each brain sample was further immersed into 50% trichloroacetic acid. After being homogenized and centrifuged, the supernatant (1 ml) of each brain sample was collected and diluted with ethanol (1:3). Finally, the signal was acquired with a multifunctional microplate reader (excitation: 620 nm; emission: 680 nm; SpectraMax M5, Molecular Devices, USA). The extravasation of EB dyes was described in micrograms/gram brain tissue weight.

| TUNEL staining
At 24 h and 72 h after ICH, the brain samples were obtained and paraffin sections were prepared. TUNEL staining was performed with an In Situ Cell Death Detection Kit (Cat. No.: 11684795910, Roche, Basel, Basel-Stadt, Switzerland) as our previous method. 45,46 The stained sections were imaged with a fluorescent microscope (Nikon, Nikon Eclipse C1, Japan), and TUNEL + cells were counted in a blinded manner.

| IHC staining
IHC staining was performed as previously reported. 44,45,55 In brief, antigen retrieval of the brain sections was executed through heat

| Western blot
Western blot (WB) was carried out as our previous method. 44

| Enzyme-linked immunosorbent assay
At 24 h after ICH, the brain samples were collected and used to detect pro-inflammatory cytokine levels using a Rat TNFα ELISA Kit TX, USA) as our previous method. 44 In brief, the prepared brain samples were put into related enzyme wells, which were pre-coated with rat TNFα, IL-6, or IL-1β antibodies and then incubated for about 1.5 h at 37°C. After being washed thrice with PBS, the brain samples were further reacted with chromogen solutions A and B. Finally, the brain samples were detected at 450 nm by using a multifunctional microplate reader (SpectraMax M5, Molecular Devices, USA).

| Statistical analysis
All data were expressed as means ± standard deviation (SD). Data analyses were conducted with SPSS 19.0 (SPSS, Inc., Chicago, IL, USA), and related diagrams were prepared with GraphPad Prism 5 (GraphPad, Inc, San Diego, CA, USA). All data were analyzed using the Shapiro-Wilk and Levene tests. If data satisfy normal distribution and homogeneity of variance, one-way analysis of variance (ANOVA) was considered, and then the least significant difference (LSD) test was applied to compare the difference among multiple experimental groups; conversely, for unsatisfied data, Dunnett's T3 test was adopted. The P-value was statistically significant when <0.05.

| PUE could alleviate ICH-induced behavioral defects and histological injury
In our experiment, a total of 341 rats were used, and five rats died

| PUE could drop ICH-induced brain cell apoptosis and hematoma volume
Typically, microscopic images of TUNEL staining were obtained at 24 h and 72 h after ICH (Figure 2A,B). Apoptotic brain cells significantly increased after ICH (P <.001, 24 h and 72 h) ( Figure 2C).  Figure 3B). Consistently, these beneficial effects of PUE at 50 mg/kg and 100 mg/kg on the hematoma size were no significant statistical difference (P > .05, 24 h and 72 h) ( Figure 3B).

| PUE could restrain ICH-induced stimulation of NF-κB pathway
Research studies have suggested that PUE can suppress the activation of the NF-κB signal pathway. [58][59][60][61] Our results also indicated that PUE (50 mg/kg) could markedly curb the upregulation of total NF-κB p65 (P < .001), p-NF-κB p65 (P < .05), and nuclear NF-κB p65 F I G U R E 2 Effects of PUE treatment at doses of 50 and 100 mg/kg on the apoptosis level of brain cells and disruption of BBB. Typical microscopic images of TUNEL + cells from the perihematomal brain tissue are shown (A, B), and relatively quantitative analyses of TUNEL + cells (C) at 24 h and 72 h after ICH induction were obtained (n = 6 rats/group; 24 h: LSD test; 72 h: Dunnett's T3 test). The quantitative analyses of extravasated EB dyes were exhibited at 24 h and 72 h post-ICH (D) (n = 6 rats/group; 24 h: LSD test; 72 h: Dunnett's T3 test). Scale bar = 50 μm. Values are presented as means ± SD. ***P < .001; **P < .01; *: P < .05 F I G U R E 3 Effects of PUE treatment at doses of 50 and 100 mg/kg on the hematoma formation at 24 h and 72 h after ICH. At 24 h and 72 h after ICH induction, typical macroscopical images of rat brains were obtained by autopsy and exhibited (A). Relatively quantitative analyses of hematoma volume were carried out and are shown (B) (n = 6 rats/group; 24 h and 72 h: Dunnett's T3 test). Values are reported as means ± SD. ***P < .001; **P < .01; *P < .05 | 7817 ZENG Et al.

| PUE could restrain ICH-induced activation of apoptosis signal
Typical WB bands of apoptosis signal-related proteins (Bcl-2, Bax, and cleaved caspase-3) are shown ( Figure 6A,B). The results showed that PUE at dosages of 50 mg/kg (P <.001) and 100 mg/ kg (P < .001) could significantly increase the expression level of Bcl-2 at 24 h after ICH ( Figure 6C). Moreover, obtained results also suggested that PUE could notably compromise the upregulation of Bax (50 mg/kg and 100 mg/kg: P < .05) and that of cleaved caspase-3 (50 mg/kg: P < .05; 100 mg/kg: P < .01) at 24 h following ICH ( Figure 6D,E). Typical IHC images of cleaved caspase-3 were acquired at 24 h after ICH and are shown ( Figure 5A).

| LY294002 could compromise PUE-mediated beneficial effects after ICH
LY294002 (a specific inhibitor of the PI3K/Akt signal) was used to ulteriorly explore the protective mechanisms of PUE in ICH-induced EBI, and the acquired results indicated that LY294002 could markedly weaken the beneficial effects of PUE on ICH-induced behavioral deficiency indicated by increasing the mNSS scores (P < .05) ( Figure 7A) and BBB disruption indexed by aggravating EB dye extravasation (P < .01) ( Figure 7B) at 24 h after ICH.

| D ISCUSS I ON
Increasing evidence has shown that OS and neuroinflammation are incredibly crucial for ICH-induced EBI, which is characterized by massive brain cell apoptosis. 1,2,5,6,62 Here, we found that PUE could efficiently alleviate ICH-induced EBI by reducing the mNSS scores, brain cell apoptosis, and hematoma volume and by improving the histological injury, BBB disruption, and brain edema. Moreover, we also found that PUE could notably repress NF-κB pathway activation and promote PI3K/Akt signal stimulation by upregulation of cytoplasmic NF-κB p65, PI3K, and p-Akt and downregulation of total, phosphorylated, and nuclear NF-κB p65. Besides, PUE could also significantly inhibit ICH-induced OS and the production of pro-inflammatory cytokines. Finally, we ulteriorly found that LY294002 could notably compromise PUE's brain beneficial effects by aggravating behavioral deficiency and BBB disruption, upregulating total and p-NF-κB p65 levels, downregulating PI3K, p-PI3K, and p-Akt levels after ICH. This was the first time to explore in detail the neuroprotective effects of PUE on ICH-induced EBI and related molecular mechanisms. The potential molecular mechanisms of PUE's brain-protective effects are shown (Fig. S6).
Research studies have confirmed that activation of NFthe -κB signal pathway widely exists in various kinds of CNS damages, including ischemic stroke, 63-65 ICH, 44,45,66 and SAH, 36,67 and its repression could notably reduce these brain impairments. 36,[63][64][65]67 In our previous research, we also have verified that NF-κB signal pathwayrelated proteins were significantly upregulated after ICH, and its suppression could markedly alleviate ICH-induced EBI and neurobehavioral deficiency. 44,45 Similarly, our current results suggest that PUE could significantly improve ICH-induced EBI and related molecular mechanisms possibly by PI3K/Akt signal activation-mediated suppression of the NF-κB pathway.
Some potential limitations deserve special attention in our study.
First, PUE can generate multiple beneficial effects by manipulation of different signal pathways, but we primarily focused on PI3K/ Akt signal activation-mediated suppression of the NF-κB pathway.
Hence, PUE may exert neuroprotective effects by other signal pathways such as the Nrf2 signaling pathway 32 and iron metabolism pathway. 25 Second, the collagenase-induced rat ICH model itself cannot completely mirror the pathophysiological process of ICH patients. Thus, results from our research might need to be further verified with other ICH models such as autologous blood and hemoglobin injection and be carefully explained.
Overall, our results have indicated that PUE could notably improve ICH-induced EBI and neurological deficiency, and related mechanisms might be involved in the suppression of NF-κB signal pathway activation-induced brain injury partly by the triggering of PI3K/Akt signal pathway-mediated neuroprotection. Our findings might provide a promising therapeutic selection for ICH-induced EBI.

ACK N OWLED G EM ENTS
This study was subsidized by the National Natural Science Foundation of China (No. 81671125).

CO N FLI C T S O F I NTE R E S T
All authors listed in this manuscript declared that no any conflicts of interest existed.

DATA AVA I L A B I L I T Y S TAT E M E N T
Contact the corresponding author for data requests.