CORM-2 can Attenuate Bleeding-mediated Inammation by Increasing Phagocytic Capacity of Cerebral Microglial Cells in Neonatal Rat in Vitro

Objective: This study aimed to explore the mechanism of CORM-2 on attenuating bleeding-related inammation. Methods: Microglia were isolated from the neonatal rats (1-2days old) and identied by the CD11b/c antibody, and some microglia were co-cultured with RBCs marked with PKH26 uorescent dye, and then treated with CORM-2. That is, the microglia cells were divided into the microglia, microglia+ PKH26+ RBCs and microglia + PKH26+ RBCs+CORM-2 cell-groups. Microglial phagocytosis to RBCs PKH26+ was observed under an inverted uorescence microscope; moreover, the uorescence intensity of microglia that phagocytized PKH26+ RBCs was detected through immunouorescence. HO-1, NF-κB p65, and IL-1β expressions were detected using RT-qPCR, western blotting, and immunouorescence, respectively. The levels of carbon monoxide hemoglobin (HbCO) in the cell supernatant in each group were detected with ELISA. phagocytosis to RBCs and decrease IL-1β and NF-κB; the mechanism may involve HO-1/CO system.


Introduction
Intracerebral hemorrhage (ICH) is one of the most fatal stroke subtypes and lacks an effective medical treatment. The predominant hematoma component is red blood cells (RBCs), which are cleaved into thrombin and hemoglobin [1,2]. Microglia clear and degrade hemoglobin into heme. In the presence of oxygen and nicotinamide adenine dinucleotide phosphate, human heme oxygenase (HO) catalyzes heme oxidation resulted in α-meso-hydroxyheme and verdoheme, and nally formed biliverdin [2]. Hematomas release numerous in ammatory factors and chemokines, which have been considered as the main factors resulting in secondary brain injury [1,3,4]. Effectively inhibiting in ammatory response in perihematoma area may reduce ICH-mediated secondary brain injury.
Microglia are the rst kind of non-neuronal cells involved in the innate immune response to ICH. In acute brain injury, microglia can be activated and polarized into two phenotypes: M1-type (pro-in ammatory) and M2-type (anti-in ammatory) microglia [5,6]. This implies microglia involvement in ICH pathological processes of both injury and repair. However, the mechanisms remain unclear, which may involve microglial production of pro-/anti-in ammatory cytokines and chemokines. Furthermore, targeted M2-likemicroglia activation may promote its phagocytosis of RBCs and tissue debris, which may bene t to hematoma clearance.
HO-1 mainly expressed in microglia/macrophages and endothelial cells post-ICH [7]. We found previously that HO-1 up-expression could improve neuro-motor function, attenuate peri-hematoma in ammatory injury, and decrease NF-κB expression in a rat brain hemorrhage model [8][9][10]. In the lipopolysaccharide (LPS)-induced in ammatory injury model, sulforaphane help microglia exerted anti-in ammation effect by increasing the expression of HO-1 in microglia [11]. Another study also revealed that HO-1/CO system was closely related to the migration, phagocytosis, and anti-in ammation ability of microglia, including HO-1 promoted CO expression, which in turn increased HO-1 expression, to active this positive feedback HO-1/CO system could promote migration and the immune phagocytic function of microglia, and decrease the in ammatory factor expression [12], which may exert protective effects in central nervous system.
CO, an endogenous antioxidant gas molecule produced by HO-1-mediated heme decomposition, exerts strong antioxidant effects. CORM-2 is a carbon-based compound that can slowly release CO, and lowlevel of CO is good for anti-in ammation and anti-microbe [13][14][15]. CORM-2 upregulates HO-1 to exert cytoprotective effect, and inhibits the activity of NF-κB [16]. However, the underlying molecular neuroprotective mechanism post-ICH of the HO-1/CO system remains unclear.
Herein, we aimed to explore the effect of HO-1/CO system on promoting microglia phagocytosis, as well as the possible mechanism of the HO-1/CO system on attenuating bleeding induced in ammation.

Microglia isolation and identi cation
Five Sprague-Dawley (SD) neonatal rats (1-2 days old) were disinfected using 75% alcohol, decapitated, and their brains removed in a sterile environment. The isolated cerebral cortex was placed in a centrifuge tube with 2 ml of Hanks solution and digested in 0.25% trypsin for 10 min. Digestion was ended with DMEM/F12 medium containing 10% fetal bovine serum (FBS); subsequently, primary mixed glial cells were obtained through ltration and centrifugation. Microglia were isolated from shaking the petri dishes for 12h, and were identi ed with CD11b/c immuno uorescence test.

Red blood cells isolation and labeled
1.0 ml of blood was collected from the left heart ventricle of 5 SD neonatal rats (0.2ml / each rat) and placed in an acid-citrate-dextrose anticoagulant solution, followed by centrifugation at 400 g × g for 5 min for RBC collection. Washing the RBCs (2 × 10 7 ) thrice with a serum-free culture medium and then centrifuging them at 400 × g for 5 min to form a loose cell mass, and removing the supernatant, nally, the RBCs were suspended in 1.0 ml of diluent to ensure complete dispersion. The RBC suspension was promptly added into the diluted dye solution (4 × 10 − 6 mol/L PKH-26), mixed them with a pipette evenly and rapidly, and incubated at 25°C for 2 to 5 min, added the same amount of FBS incubated for 1 min to stop the staining reaction, then centrifuged at 400 × g for 10 min, removed the supernatant, the labeled RBCs were washed thrice with serum-containing medium to obtain PKH26+ RBCs.

Co-culture of RBCs and microglia
Microglia and PKH26+ RBCs were mixed at a 1:10 ratio and co-cultured in 37°C with 5% CO 2 for 1day, meanwhile, some of them were treated with CORM-2 (50 Um), to get the followings cell-groups: Microglia, Microglia + PKH26+ RBCs, and Microglia + PKH26+ RBCs + CORM-2. At the 6th h and the 24th h and the 3rd day after co-cultivation, respectively, the phenomenon of microglial phagocytosis to the PKH26+ RBCs would be observed under the inverted uorescence microscope.

Immuno uorescence
After fusion to 80%-90%, cells in the cell-groups mentioned above were harvested and washed thrice with phosphate-buffered saline (PBS). The cells were xed with 4% paraformaldehyde for 15 min and washed thrice with PBS. Next, 0.5% Triton X-100 was added for 20 min. After 30 min of blocking with 5% bovine serum albumin at 37°C, the cells were incubated with uorescein isothiocyanate-conjugated primary antibody at 4°C overnight. Cy3-conjugated secondary antibody diluted with blocking buffer was used for staining at 37°C for 30 min. Next, the cells were incubated with DAPI in the dark for 5 min. The samples were sealed with 50% glycerol and observed under a uorescence microscope. The expression of HO-1, NF-Κb p65, and IL-1βin all cell-groups at the 1st, 3rd, and 5th day were compared with RT-PCR, Western blot, ELISA, respectively.

RT-qPCR
Total RNA was extracted using an Ultrapure RNA kit (Takara, Tokyo, Japan) according to the manufacturer's protocol and subjected to reverse transcription using a HiFiScript cDNA Synthesis Kit. Real-time PCR was performed using SYBR Green PCR Master Mix. The β-actin gene was used as an internal control to normalize RNA quantity and quality differences in all samples. Target gene mRNA quanti cation was performed using the 2 −△△ Ct method. Table 1 shows the primer sequences.

Western blot
Protein extraction was performed for the different cell groups. Brie y, the cells were harvested and lysed, followed by protein extraction. Proteins were quanti ed using the BCA kit and loaded in SDS-PAGE gel (12%, 5-10 µg per lane); subsequently, they were electrotransferred to a PVDF membrane. The membrane was rinsed and placed in blocking buffer. Samples were incubated at 4°C overnight after adding the primary antibodies. Next, the membrane was rinsed and incubated with a secondary antibody at room temperature for 2 h. Photoelectrons were captured using the ImageQuant LAS 4000 imaging station (GE) followed by quanti cation.

ELISA
The levels of HbCO expression were determined with ELISA kit according to the manufacturer's instructions. The kit employs a double-antibody sandwich method. Puri ed rat HbCO was used for antibody capturing and coated with a microporous plate to make a solid-phase antibody. The cell supernatant was sequentially added to the coated microporous, followed by the addition of the horseradish peroxidase-labeled detection antibody to form an antibody-antigen-enzyme antibody complex. The absorbance (optical density value) was measured at a 450-nm wavelength using a microplate reader; moreover, the HbCO level in the sample was calculated from the standard curve.

Statistical analysis
Statistical analyses were performed using SPSS 22.0(IBM SPSS 22.0, SPSS Inc). All data were presented as the mean ± standard deviation (SD). Between-group differences were evaluated using a one-way analysis of variance. Statistical signi cance was set at P < 0.05.

Microglia identi cation
To identify microglia among the isolated primary cells, CD11b/c expression was detected using immuno uorescence. As shown in Fig. 1, cells with positive CD11b/c (Green uorescence) were identi ed as microglia.

The effect of CORM-2 on microglia phagocytosis
After 6 h of PKH26+ RBCs (small cells with red uorescence) and microglia co-culture, the phenomenon of PKH26+ RBCs were wrapped in microglia (large cells with red uorescence) were found under a 20-fold inverted uorescence microscope. The number of PKH26+ RBCs (small red spots, white arrow) and the number of microglia who swallowed PKH26+ RBCs (large red spots, red arrow) in the co-culture model were calculated, respectively, using a blood-cell counting board at the 24th h and the 72nd h after CORM-2 treatment.
We noticed that the number of microglia who swallowed PKH26+ RBCs (Red arrow) were gradually increasing from 24 h through 72 h after co-culture in both microglia + PKH26+ RBCs group and microglia + PKH26+ RBCs + CORM-2 group. Which indicated that more and more RBCs were swallowed by microglia in all the two cell-groups over time.
More importantly, we found that the number of microglia who swallowed PKH26+ RBCs (Red arrow) increased more remarkably in CORM-2 group than that in non-CORM-2 group at both the 24th h and the 72nd h, meanwhile, the number of PKH26+ RBCs (White arrow) decreased more remarkably in CORM-2 group than that in non-CORM-2 group accordingly at the same time points (Fig. 2), all of which indicated that CORM-2 increased the microglia phagocytosis to RBCs.
3. The effect of CORM-2 on HO-1, NF-κB p65, and IL-1β expression RT-qPCR revealed that at the 1st day, the levels of HO-1 in all cell-groups were as high as the same, without showed statistic difference, while the levels of IL-1β and NF-κBp65 in microglia + PKH26+ RBCs, and microglia + PKH26+ RBCs + CORM-2 groups were remarkably higher than that in microglia group, which revealed that RBCs induced in ammation. However, the levels of IL-1β and NF-κBp65 in microglia + PKH26 + RBCs group were higher than that in microglia + PKH26 + RBCs + CORM-2 group, which revealed that CORM-2 can inhibit in ammation and consume HO-1. At the 3rd d, HO-1 mostly increased in microglia + PKH26+ RBCs + CORM-2 group, and decreased in microglia + PKH26+ RBCs group, all of which revealed that CORM-2 increased HO-1 expression. IL-1β and NF-κB decreased remarkably at the 3st day and the 5th day in microglia + PKH26+ RBCs + CORM-2 group (Fig. 3).
Western blot analysis revealed that at the 1st day, HO-1 protein expression showed no signi cant differences among all cell-groups; whereas, at the 3rd day and the 5th day, the HO-1 protein expression was most remarkably the microglia + PKH26+ RBCs + CORM-2 group, which indicated that CORM-2 signi cantly increased HO-1 protein expression. At the same time, CORM-2 signi cantly reduced IL-1β and NF-κB protein expression (Fig. 4).
Immuno uorescence showed no signi cant differences of HO-1 expression on the 1st day and the 3rd day. On the 5th day, higher level of HO-1 expression was found in microglia + PKH26+ RBCs + CORM-2 group; and the lower levels of IL-1β and NF-κB p65 were also in microglia + PKH26+ RBCs + CORM-2 group, when compared with other two groups (Fig. 5). Which revealed that CORM-2 could increase HO-1 and decrease L-1β and NF-κB p65 expression.

4.HbCO levels in the co-culture model
As shown in Fig. 6, there was no signi cant difference of HbCO levels among cell-groups at the 1st day.

Discussion
ICH-induced brain injury can be divided into two phases: primary injury caused by the mass effect of intraparenchymal hematoma and secondary injury caused by oxidative stress and neuroin ammation in the perihematomal area [17,18].
Hematoma and its degradation products may activate post-ICH in ammatory responses in the perihematomal region [19]. Effective hematoma removal is crucial for regulating in ammation and functional recovery [20]. Microglia are crucial for tissue repair involving hematoma and damaged-cell phagocytosis post-ICH; moreover, they exert anti-in ammatory and pro-in ammatory effects. Microglia are innate immune cells in the brain and are considered as the rst kind of non-neuronal cells to respond to various acute brain injuries, including ICH [21,22]. Therefore, this study established a microglia/RBC co-culture model in vitro to explore the microglial involvement in ICH.
Hematoma removal and absorption are essential for ICH recovery and are clinically achieved through craniotomy or minimally invasive hematoma removal surgery and drug treatment. Moreover, the quality of clinical outcomes is positively correlated with the speed of hematoma absorption [23]. Therefore, promoting endogenous hematoma absorption has become a novel ICH treatment strategy [24]. Microglia, which are effector cells of immune and in ammatory responses in the central nervous system, can clear hematomas and protect nerve cells by phagocytizing RBCs and dissolved RBC components. In the LPSinduced in ammatory injury model, the HO-1/CO system was found to promote microglia migration, accelerate microglial phagocytosis, and protect central nervous system [25]. In the microglia/erythrocyte co-culture model in this study, the microglia + PKH26+ RBCs + CORM-2 group was con rmed have more strong e cacy on increasing HO-1and decreasing IL-1β and NF-κB p65. All of which indicated that CORM-2 could increase microglial phagocytosis of RBCs and inhibit bleeding-induced in ammation.
In addition to accelerating hematoma absorption, it is important to regulate the in ammatory response around the hematoma. Neuronal injury and related neurological outcomes are dependent on a delicate balance between pro-in ammatory and anti-in ammatory mediators [26]. Endogenous CO is mainly oxidized by heme, widely involved in cardiovascular diseases, respiratory lesions, and other physiological /pathophysiological processes, has been con rmed with anti-in ammatory, anti-apoptotic, and antioxidative effects [13,27].
CO was con rmed could inhibit the expression of in ammatory factors, such as IL-1β, and macrophage in ammatory protein 1β, and reduce in ammatory injury [28,29]. The HO-1/CO system can inhibit TNF-αmediated in ammatory response [27], and CORM-2 can not only up-regulate HO-1 but also induce the cytoprotective effect of HO-1 [30]. This study revealed the mechanism underlying erythrophagocytosis modulation by the HO-1/CO system and neuroprotection by microglia. Activated microglia can release reactive oxygen species, which cause protein oxidation, membrane lipid peroxidation, enzyme inactivation, and DNA damage [31]. However, Mayne reported that decreased microglial TNF-α expression reduced neuronal apoptosis around the hematoma and improved the neurobehavioral score [32]. Our ndings con rmed that CORM-2 effectively increased HO-1 expression, as well as inhibited NF-κB p65 and IL-1β expression in the microglia and RBC co-culture model, which suggested that CORM-2 could induce the microglia anti-in ammatory effect. However, it remains unclear whether CORM-2 inhibits the NF-κB signaling pathway by suppressing NF-κB p65 subunit phosphorylation or the nuclear translocation process of the NF-κB p65 subunit. HO-1 could inhibit the production of numerous downstream in ammatory factors of NF-κB by inhibiting and promoting NF-κB p65 and nuclear factor 2-related factor (Nrf2) entry, respectively, into the nucleus. Therefore, it plays a neuroprotective effect on early injury around the cerebral hemorrhage focus in rats [8][9][10]. In this study, CORM-2 signi cantly inhibited cellular and nuclear NF-κB p65 expression, which indicates that the HO-1/CO system may enhance the microglia anti-in ammatory effect by inhibiting post-ICH nuclear translocation of NF-κB p65 subunit.
Finally, we assessed whether CORM2 increased HbCO and affected oxygen metabolism. The a nity of carbon monoxide to hemoglobin is approximately 200 times greater than that of oxygen; therefore, CO poisoning could cause hypoxic damage. CORM-2 can slowly release CO in DMSO solution, which is convenient for controlling the CO release rate [33]. We found that CORM-2 adding to the microglia + RBC co-culture model did not increase the HbCO levels; instead, it mildly decreased HbCO. CORM-2 could enhance microglial phagocytosis to RBCs, which decreased hemoglobin level. In this study, although there was a gradual increase in CO expression, HbCO saturation remained stable (500-600 ng/ml) after CORM-2 treatment, indicated that CORM-2 did not produce serious toxic and side effects. And also, CORM-2 did not result in excessive carboxyhemoglobin levels. Study identi ed that low level of HbCO did not insigni cantly affect overall mitochondrial function and biogenetics, but resulted in a signi cant increasing in the basal oxygen consumption rate. Assessment of mitochondrial function with inhibitors revealed no other alterations in the oxygen consumption rate [34]. Although CO gas has already passed safety evaluation in phase I trials in healthy humans and possesses a backbone carrier moiety, CORM-2 should be stringently characterized from a metabolic and toxicological perspective [35]. Further study is needed to elucidate the pharmacokinetics and biology of CO and CORMs.

Conclusions
CORM-2 can inhibit in ammatory reactions in bleeding setting in vitro by promoting microglial phagocytosis to RBCs and decrease IL-1β and NF-κB; the mechanism may involve HO-1/CO system.

Declarations
Funding Figure 1 Identi cation of microglia Microglia and PKH26+RBCs RBC density in the co-cultures White arrow indicates the PKH26+RBCs, Red arrow indicates the microglia swallowed RBCs.