Carbon monoxide attenuates vasospasm and improves neurobehavioral function after subarachnoid hemorrhage.

Abstract

Subarachnoid hemorrhage (SAH) is a devastating form of hemorrhagic stroke and is a serious medical condition caused by bleeding usually due to a ruptured aneurysm. Oxidative stress and inflammation from hemoglobin and heme released from lysed red blood cells are some postulated causes of vasospasm during SAH, which could lead to delayed cerebral ischemia. At low amounts, carbon monoxide (CO) gas may be neuroprotective through anti-inflammation, anti-cell death, and restoration of normal blood flow. Hence, this study focuses on a noninvasive strategy to treat SAH by using CO as therapeutic medical gas. Mice were treated with 250ppm CO or air for 1h started at 2h after SAH. Various anatomical and functional outcomes were monitored at day 1 and day 7 after SAH. CO decreased neurological deficit score (47.4±10.5%), and increased activity (30.0±9.1%) and stereotypic counts (261.5±62.1%) at 7d. There was a significant increase in lumen area/wall thickness ratio in the middle cerebral artery (173.5±19.3%) and tended to increase in the anterior cerebral artery (25.5±4.3%) at 7d. This is a first report to demonstrate that CO minimizes delayed SAH-induced neurobehavioral deficits, which suggest that post-treatment with CO gas or CO-donors can be further tested as a potential therapy against SAH.

1. Introduction

Subarachnoid hemorrhage (SAH), is a type of hemorrhagic stroke that involves a major complication of bleeding in the subarachnoid space because of an aneurysmal rupture [1–3]. Bleeding in the brain is a common complication that occurs after an aneurysmal accident in SAH leading to brain injury [4–6]. In addition, SAH induces vasospasm, which leads to neuroinflammation, oxidative stress and neuronal injury and delayed cerebral ischemia; and it is also considered a major cause of mortality and morbidity [7]. Delayed neurologic deterioration from vasospasm remains the greatest cause of death and disability after SAH [8,9]. Cerebral vasospasms occur in more than one-half of all patients with SAH. and it is recognized as the main cause of delayed cerebral ischemia after SAH [10–12]. This condition normally advances between 4 and 14 days and most severe conditions are observed between 7 to 9 days after SAH [13]

Carbon monoxide (CO) has a potential impact on pathophysiological conditions such as neuronal death, neuroinflammation, cell metabolism, neuromodulation and vasomodulation [14–16]. CO is produced endogenously through the enzyme heme oxygenase (HO) as it breakdowns heme, an important pro-oxidant generated due to hemolysis. Studies have also concluded that most of the endogenously generated CO is HO-dependent [17,18]. Other evidence shows that CO in the body is also formed by photo-oxidation, lipid peroxidation, and xenobiotic metabolism [19]. We have previously shown that the constitutive HO2 and inducible HO1 have a protective role in the stroke brain pathology [20–23]. Furthermore, studies have also concluded that HO-derived CO, CO-releasing molecules [24,25], and low concentrations of inhaled CO protect against various insults through the activation of anti-inflammatory, anti-cell death, and vasodilatory effects [26,27]. While in most scenarios the free heme substrate can be the limiting factor for the production of large amounts of CO, our hypothesis is that exogenous CO might provide protection against SAH-induced vasospasm and neurobehavioral deficits. The levels of CO examined in this study are safe and these levels are usually reported within the physiologically environment. Here we designed an observational study to examine the effect of CO inhalation, compared to air delivered at the same flow rate on functional outcomes in an endoperforation model of SAH.

We report for the first time that CO is neuroprotective in endoperforation model of SAH. This study will provide the initiative to investigate the neuroprotective mechanisms of CO in SAH and whether CO or CO-donors can be used as potential therapeutic candidates for SAH treatment.

2. Materials and Methods

2.1. Animals and induction of SAH by endoperforation

Eight to ten week old male wild-type (C57BL/6), mice were used to induce SAH t [28]. Mice were anesthetized and common carotid (CCA), internal carotid (ICA) and external carotid (ECA) arteries were exposed. A 5–0 nylon monofilament was passed into the ICA though the ECA and extend until a slight resistance was felt. At that point, the filament was advanced further that caused perforation of the blood vessel along the circle of Willis. Sham control mice were subjected to the same surgical procedure without endoperforation. All postoperative care was given by the experimenter and UF veterinarian technicians. All experiments were approved by the University of Florida Institutional Animal Care and Use Committee recommendations and followed the ARRIVE guidelines [29]. All personnel performing the surgery, functional assessments, histology, and histological assessments were blinded to the experimental conditions.

2.2. CO treatment in SAH-induced mice

We used one time point for CO exposure that started immediately at 2h after SAH on day 1. Mice were exposed to CO (SAH+CO) or medical grade air (SAH+Air) once daily for 1h until 7d. Mice were placed inside a Plexiglas chamber at room temperature for exposure with 250ppm of CO or Air at the same flow rate of 1L/min. We adopted the dose of CO and treatment time point from our previous study [30]. CO level in the chamber was monitored by a Single Gas Analyzer, (CO91, Universal Enterprises, Beaverton, OR). Sham control (Sham+Air) mice received air once daily for 1h until 7d at the same flow rate. After 1h of CO or air exposure, mice were removed from the chamber and placed in their home cages.

After the surgery, mice were assessed for neurobehavioral functions, such as motor function, locomotor activity, and neurological deficit at 24h and 7d. After the behavioral assessments, mice were perfused and the brains were harvested and kept at −80°C until processing. Ten micrometers of brain sections were obtained using a cryostat and were stained with hematoxylin and eosin to analyze vasospasm in the middle cerebral artery (MCA) and anterior cerebral artery (ACA). Representative experimental design is shown in figure 1.

Fig. 1.

Schematic representation of experimental procedure: At 2h after SAH mice were exposed to 250 ppm CO or air for 1h followed by daily exposure for 1h until day 7. Mice were tested for functional outcomes at day 1 and day 7. On day 7 after functional assessments, mice were perfusion fixed and brains were harvested for vasospasm studies.

2.3. Functional assessments

Functional outcomes after SAH were tested by an experimenter blinded to the experimental conditions. For open field activity and rotarod testing, each group received pre-testing before surgery followed by assessments at 24h and 7d after SAH. All the functional assessments were performed at the same time of the day and in between each assessment 30 to 45 min of rest was given to the mice.

2.3.1. Neurological deficit score (NDS)

We used a 24-point scoring system originally introduced by Clark et al. [31] and previously reported by us to evaluate neurological deficit [32,33]. This scoring system includes body symmetry, gait, climbing, circling behavior, front limb symmetry, and compulsory circling. Each test score is graded from 0 to 4, thus establishing a maximum deficit of 24 points.

2.3.2. Open field locomotor activity test

This test is a sensitive method used to assess gross and fine locomotor activity including ambulatory and stereotypic counts. Open field activity was monitored with an automated MED Associates (Med Associates, Inc., St. Albans, VT) video tracking interface system following the procedure we have reported previously [32]. Mice were pretested for 3d before surgery to obtain a baseline activity. Thereafter, these activities were observed 1 and 7d after SAH. In brief, mice were individually housed in four transparent acrylic cages and their simultaneous activities were recorded for a period of 30 min.

2.3.3. Rotarod test

This test is used for the analysis of motor coordination and sensorimotor function. Mice were pretested for 3d before surgery to obtain their baseline functions. Thereafter, mice were tested on 1 and 7d after SAH. Mice were placed on a rotating beam that started with 5 rpm and gradually accelerated at fixed interval to reach the maximum speed of 5 to 30 rpm [21,32]. Each trial had a maximum duration of 5 min. Each group of mice underwent three trials with a 30 min interval. The total time of retention on the rod or latency to fall for each mouse was automatically recorded by Rotamex computer software (Columbus Instrument, Columbus, OH).

2.4. Mortality

Mortality rate is a measure of the number of animals that died during the course of experiment. We observed mortality throughout the experiment up to 7d after SAH.

2.5. Hematoxylin and eosin staining and assessment of vasospasm

At the terminal end point after functional assessments (7d after SAH), mice were deeply anesthetized with isoflurane and transcardially perfused with phosphate buffered saline and 4% paraformaldehyde. The gravity perfusion method was used to perfuse and fix the mice to prevent the possibility of vessels deformity. Brains were carefully removed and post-fixed in 4% paraformaldehyde for 24h. After that, brains were transferred to 30% sucrose until they settled down at the bottom of the tube. Thereafter, the brains were placed in a small plastic container that was filled with optimal cutting temperature compound and then the container was dipped in chilled 2-methyl butane over dry ice to snap freeze the brain. This technique was used to keep the MCA and ACA intact during cryosectioning. Thereafter, brains were cut into 10 μm sections on a cryostat, and stained with hematoxylin and eosin. Representative digital images of eight consecutive MCA and ACA cross sections from each animal, were used quantify the lumen area/wall thickness ratio and the lumen circumference/wall thickness ratio to assess vasospasm [34,35]. All slides were scanned using an Aperio ScanScope CS and analyzed with Image Scope software (Leica Biosystems, Cincinnati, OH).

2.6. Statistical analysis

Statistical analyses were performed by using GraphPad prism 5. A student’s t-test was used to determine significant differences between SAH+Air and SAH+CO groups in Fig. 4. For the rest of the data, one way analysis of variance (ANOVA) followed by post-hoc Tukey’s multiple comparison tests was performed to estimate statistical significance among the groups. Results are considered statistically significant at p<0.05. Mean value of data also represented with 95% of confidence interval (CI) and with two-tail p value. All data are presented as mean±SEM.

Fig. 4.

Effect of CO on vasospasm: Representative microscopic images showing right MCA or right ACA (A). There was significantly improved lumen area/wall thickness in MCA region (B); and lumen circumference/wall thickness ratio in MCA and ACA at 7d in CO-treated group as compared to Air-treated group (C). There was also an increase in lumen area (D) and lumen circumference (E) and; decreased wall thickness (F) by CO-treatment in SAH mice at 7d in MCA region. However, no percent difference was observed in lumen area/wall thickness in ACA region, though wall thickness was significantly reduced and lumen area was significantly increased by CO treatment as compare to SAH at 7d (n=4–6). Data presented as mean±SEM. *p<0.05, **p<0.01, ***p<0.001 as compared with SAH+Air.

3. Results

3.1. SAH induces vasospasm

To study the time dependent changes in vasospasm, we performed 24h, 3d, 5d and 7d survival studies after SAH (Fig. 2). Representative brain images showing the hematoma in SAH mice brain (Fig. 2A). Microscopic images showing hematoxylin and eosin stained cross section of ACA from different groups (Fig. 2B). We determined vasospasm by finding the vessel lumen area/wall thickness ratio and presented in percent difference. Interestingly, we found significant vasospasm by comparing the lumen area/wall thickness ratio in mice brain after SAH at 24h (313.3±45.7, p<0.001, CI: 219.2.0 to 407.4), 3d (497.0±110.0, p<0.001, CI: 268.9 to 725.2), 5d (918.3±118.0, p<0.001, CI: 674.1 to 1162.0), and 7d (1508.0±180.5, p<0.05, CI: 1137.0 to 1879.0) compared to sham (2078.0±133.2, CI: 1802.0 to 2353.0) (Fig. 2C). We also separately measured the lumen area and wall thickness to correlate any significant changes in the lumen wall and lumen area and we expressed the results as the percent difference as compared with the sham. The lumen area at 24h was 25.6±2.9% (p<0.001, CI: 19.5 to 31.7), 3d was 38.6±8.0% (p<0.001, CI: 21.9 to 55.3), 5d was 64.0±6.1% (p<0.001, CI: 51.4 to 76.7), and 7d was 78.1±5.9% (p<0.05, CI: 66.0 to 90.2) as compared to sham 100.0±4.1% (CI: 91.5±108.5) (Fig. 2D). Wall thickness also changed and at 24h it was 178.6±9.1% (p<0.001, CI: 159.9 to 197.7), at 3d it was 182.2±10.9% (p<0.001, CI: 159.5 to 205.0), at 5d it was 167.0±10.8% (p<0.001, CI: 144.6 to 189.4), and at 7d it was 119.5±6.7% (p>0.05, CI: 105.7 to 133.4) as compared to 100.0±3.8% in sham (Fig. 2E).

Fig. 2.

Time dependent changes in vasospasm after SAH: Macroscopic images showing brains after SAH (A). Representative images of the brain section after hematoxylin-eosin staining showing ACA (B). There was significant less lumen area/wall thickness ratio at 24h, 3d, 5d, and 7d in comparison to sham for the right ACA (C). There was also significant reduction in lumen area (D) and wall thickness (E) at 7d in SAH group as compared to sham control (n=3–4). There was a significant increase in NDS score at 24h, 3d, 5d and 7d as compared to sham after SAH (n=5–6) (F). Data presented as mean±SEM. *p<0.05, **p<0.01, ***p<0.001 as compared with sham.

3.2. Effect of SAH on neurological deficit

Mice were tested for NDS at different time points after SAH. SAH induced mice exhibited significantly higher NDS at 24h (12.0±0.3; p<0.001; CI: 11.2 to 12.9), and on days 3 (12.0±0.4, p<0.001, CI: 10.8 to 13.5), 5 (12.2±0.8, p<0.001, CI: 10.0 to 14.4), and 7 (11.7±0.6, p<0.001, CI: 10.1 to 13.3) as compared to sham (2.3±0.3, CI: 0.9 to 3.8) (Fig. 2F).

3.3. Mortality rate

No mortality was observed in the sham group (Fig. 3A). There was 37.5% (day 1, 25% and day 2, 12.5%) mortality in the SAH+Air group and 40% mortality in the SAH+CO group (Fig. 3A). To analyze the mortality rate, the Log-rank (Mantel-Cox) Test (Chi square=1.678, p<0.4322) was performed. There was no significant difference observed in mortality rate between SAH+CO and SAH+Air treated groups. This range of mortality was similar to that others using the endoperforation model of SAH [28,36].

Fig. 3.

Animal mortality curve and hematoma observation: Animal mortality curve showed that there was 37.5% (25% on day 1 and 12.5% on day 2) mortality in SAH+Air-treated group whereas there was 40% mortality reported on day 1 in SAH+CO group (3A). SAH brain at 7d did not shows any difference or sign of hematoma in SAH and SAH+CO-treated mice brain (3B).

3.4. Assessment of hematoma or hemorrhage at 7 day

Following morphological observation of the brains after perfusion, no sign of hemorrhage or hematoma was observed at 7d; accordingly, no differences were observed between SAH+Air- and SAH+CO-treated groups (Fig. 3B).

3.5. CO treatment attenuates SAH-induced vasospasm

We induced SAH by endoperforation along the circle of Willis in wildtype mice to study vasospasm. After 2h of SAH, we exposed the mice to 250ppm CO or air for 1h followed by a single exposure to CO or air for 1h daily for 7d. Interestingly, we found significantly improved lumen area/wall thickness ratio and in the ratio at 7d in the SAH+CO group compared to the SAH+Air group. CO significantly decreased vasospasm as noted by a higher lumen area/wall thickness ratio in the right MCA in SAH+CO-treated group (944.8±66.7, p<0.001, CI: 808.1 to 1081.0) to SAH+Air-treated group (345.4±55.3, CI: 232.1 to 458.7) at 7d. We observed a trend towards a higher lumen area/wall thickness ratio in the right ACA in the SAH+CO-treated group (2073.0±117.5, p>0.05, CI: 1833.0 to 2312.0) compared to SAH+Air-treated group (1918.0±246.8, CI: 1417.0 to 2420.0) at 7d (Fig. 4A–B).

We also analyzed vasospasm by measuring the lumen circumference/wall thickness ratio. CO significantly decreased the vasospasm as noted by a higher lumen circumference/wall thickness ratio in the right MCA at 7d (SAH+Air, 8.3±1.9, CI: 4.4 to 12.2 vs SAH+CO, 18.7±2.4, p<0.01, CI: 13.7 to 23.7) and in the right ACA at 7d (SAH+Air, 15.4±2.3, CI: 10.6 to 20.2 vs SAH+CO, 20.7±1.2, p<0.05 CI: 18.3 to 23.1) after SAH (Fig. 4C).

And, also measured the lumen area, lumen circumference and wall thickness separately to correlate any significant changes in the lumen wall thickness and lumen area and we expressed the results as the percent difference compared to the SAH+Air group. In MCA we found a significantly higher lumen area in the SAH+CO-treated group (139.7±8.4%, p<0.05, CI: 122.1 to 156.4) as compared to the SAH+Air-treated group (100.0±15.8%, CI: 67.6 to 132.4) (Fig. 4D). We also observed significantly smaller wall thickness in the SAH+CO (50.9±1.6%, p<0.001, CI: 47.5 to 54.1) group as compared to SAH+Air-treated mice (100.0±4.0, CI: 91.5 to 108.7) in MCA. Similarly in ACA, we found a significant change in the lumen area in the SAH+CO-treated group (189.7±8.4%, p<0.001, CI: 168.5 to 210.9) as compared to the SAH+Air group (100.0±11.0%, CI: 78.9 to 123.6) (Fig. 4D). Lumen circumference was also measured and expressed the results as the percent difference compared with the SAH+Air mice. We found a significantly higher lumen circumference (100.0±18.9%, CI: 83.0 to 157.9 vs 166.7±10.0%, p<0.05, CI: 146.2 to 187.1) in the MCA region in the SAH+CO group as compared to SAH+Air group. However, no significant change in lumen circumference (SAH+Air, 100.0±11.2%, CI: 77.1 to 123.0 vs SAH+CO, 91.7±5.0%, p>0.5, CI: 81.5 to 101.9) in ACA (Fig. 4E). There was also a significant decrease in wall thickness in the SAH+CO-treated group (68.7±2.4%, p<0.001, CI: 63.9 to 73.5) compared to the SAH+Air group (100.0±3.7%, CI: 92.5 to 107.5) in ACA (Fig. 4F).

3.6. CO treatment attenuates SAH-induced neurological deficit

We observed NDS in the different groups of mice at day 1 and 7 after SAH. We found that SAH in mice caused significant neurological deficits at days 1 and 7; 17.0±1.8 (p<0.001, CI: 11.3 to 22.7) vs 7d; 9.5±0.6 (p<0.01, CI: 7.4 to 11.5). Interestingly, SAH+CO group exhibited significantly improved NDS 5.7±1.0 (p<0.05, CI: 2.5 to 9.0) as compared to the air treatment 9.5±0.6 (CI: 7.5 to 11.3) at day 7. However, no significant changes were observed between the SAH+Air and SAH+CO groups at day 1 (Fig. 5A).

Fig. 5.

Effect of CO on functional outcomes after SAH: There was significant increase in NDS at day 1 and day 7 in SAH+Air and SAH+CO groups as compared to sham. However no significant difference observed in NDS between SAH+Air and SAH+CO-treated group at 1 day. However, treatment with CO significantly decreased the NDS score in SAH mice at 7d (n=5–6) (A). In open field activity, a significant decrease in ambulatory distance was found in SAH+Air and SAH+CO-treated groups in comparison to sham on day 1 and day 7. However, CO treatment significantly improves ambulatory distance at 7d but no difference observed at 1 day (B). Additionally, there was significant decrease in stereotypic counts in SAH+Air mice at 1 day and 7d; whereas, CO treatment significantly improved the stereotypic counts at day 1 and day 7 (n=4–5) (C). Significant motor function deficit (latency of fall from rod) was found in the SAH+Air- and SAH+CO-treated groups in comparison to sham. However, no significantly change in latency to fall was observed in CO-treated group as compared to Air-treated group at either 1 day or 7d (n=5–6) (D). Data presented as mean±SEM. *p<0.05, **p<0.01, ***p<0.001 as compared with sham; ^#p<0.05 and ^##p<0.01 as compared to SAH+Air vs SAH+CO on day 7.

3.7. CO treatment improves ambulatory distance and stereotypic counts in SAH

We measured ambulatory distance using an automated open field activity monitor and a video tracking interface system (Med Associate). We found that SAH led to a significantly smaller distance travelled as compared to sham on day 1 (Fig. 5B). However, there was no significant difference in ambulatory distance between the SAH+Air (236.2±41.2, CI: 138.9 to 696.0; p>0.05) and SAH+CO (271.7±39.9, p>0.05, CI: 86.9 to 723.5) groups. Interestingly, we observed a significant difference ambulatory distance between SAH+Air, 499.4±37.2, CI: 381.1 to 617.7 vs SAH+CO, 649.2±45.7 (p<0.05, CI: 503.8 to 794.5) on day 7 (Fig. 5B).

We also found that SAH led to significantly lower stereotypic counts as compared to sham on day 1 (Fig. 5C). In addition, we also found that CO treatment (2458.0±422.5, p<0.05, CI: 1285.0 to 3631.0) significantly improved the stereotypic counts in SAH mice at 7d as compared to SAH+Air (680.0±231.0, CI: 38.7 to 1321.0) (Fig. 5C).

3.8. Effect of CO treatment on motor function after SAH

When mice were subjected to the rotarod test, we found that there was a significant change in motor function (latency to fall from a rotating rotarod) in SAH+Air and SAH+CO mice compared to sham groups. However, latency to fall did not improve with CO treatment in SAH mice on day 1 (54.9±21.2, p>0.05, CI: 4.7 to 105.2) or 7 (92.2±33.7, p>0.05, CI: 12.6 to 171.8), although a trend of lower latency to fall was observed in SAH+CO on day 7 (Fig. 5D).

4. Discussion

This study was designed to first characterize the time dependent changes in vasospasm in an endoperforation model of SAH. Significant increase in vasospasm at 24h, 3d, 5d, and 7d as evaluated by lumen area/wall thickness ratio was found. Aperio image scope was used to assess the lumen area and wall thickness to measure vasospasm and such method would correct the potential issues due to vessel deformity. Then, we evaluated the effect of CO treatment on SAH-induced vasospasm, neurological deficit, ambulatory distance, stereotypic counts and motor function. We found that there were significant benefits offered by CO in minimizing SAH-induced delayed vasospasm/vasoconstriction. Additionally, CO also improved neurological deficit and ambulatory function. While, CO has been shown to have benefits in other acute brain injury models, this would be the first report to document that CO prevents SAH-induced neurological deficit, locomotor, motor function, and vasospasm.

SAH is caused by bleeding in the subarachnoid space of the brain and is considered the most life threatening type of stroke due to the high mortality and morbidity associated with it [7,37]. Reports have established that neurobehavioral alterations and vasospasms are serious complications after SAH [13,38,39]. Low doses of exogenous and endogenous physiological concentrations of CO have shown protection against various types of stroke and hypoxia [30,40,41] but have not yet been tested in an endoperforation model of SAH., as tested here. The endogenous bioactive gas CO is produced all over the body, including the brain, by HO enzymes in response to various stresses and physiological conditions. There is still active debate regarding as to where the free heme pool would be accessible to produce – when needed – sufficient molar concentrations of CO that would match its exogenous-associated beneficial effects. However, determining how the exogenous supply of relatively low levels of CO will improve neurological function and vasospasm after SAH highly clinically relevant.

Several experimental models are used to recapitulate the SAH pathology and here, we used the endoperforation model of SAH as it mimics some clinical features of SAH [42]. However, a limitation of this model is that the amount of bleeding after SAH may be challenging to control; consequently, the magnitude of vasospasm is proportional to the amount of bleeding after SAH. The effects of CO after SAH are not well explained or well-described in mouse preclinical endoperforation SAH models, although they are reported in a blood injection model [23,43]. Cerebral vasospasm is considered as an important cause of morbidity in patients after SAH, where bleeding results in the generation of various cells, bioactive molecules and toxins in the nervous system that promotes neuronal injury. Therefore, we diligently looked at the time dependent effects of SAH on vasospasm in mice by analyzing blood vessel lumen area and thickness. We first investigated vasospasm in mice at different time points and found that vasospasm was consistent after 24h up to 7d. We also found significant neurological deficits at 7d compared to sham controls.

We then tested the effect of low level of CO (250ppm) on vasospasm and neurological function. In our previous studies we have reported that such low level of exogenous CO keep levels of carboxyhemoglobin within the safe range. We choose the 7d time point because maximum vasospasm is observed at 5–7d in clinical settings [13]. In a separate cohort, after the confirmation of the SAH time point, we investigated the effect of CO on SAH-induced vasospasm and neurological function by using three different neurobehavioral tests with a high sensitivity for detecting deficits after SAH. Our laboratory and others have used NDS successfully in ischemic and hemorrhagic stroke pathology [22,31,32]. The present study demonstrates that there was significant impairment in motor function after SAH. In addition, we also monitored animal activity by measuring ambulatory distance and stereotypic counts and found a significant decrease in ambulatory distance and stereotypic counts after SAH. Interestingly, CO treatment significantly improves motor function and locomotor activity.

Cerebral vasospasm after aneurysmal SAH is a well-established phenomenon that is caused by narrowing of large and medium-sized intracranial arteries [44]. Clinically, vasospasm affects the anterior circulation that is supplied by the internal carotid arteries, and vasospasm after SAH causes delayed cerebral ischemia, which continues to be a major complication and source of morbidity in cases of aneurysmal SAH [45,46]. Furthermore, we have demonstrated the effect of CO on vasospasm by estimating lumen area, lumen circumference, wall thickness, and the ratio of lumen area to wall thickness and lumen circumference to wall thickness to confirm the degree of vasospasm. Although, lumen area to wall thickness ratio is one of the way to show vasospasm, we have also measured vasospasm by a lumen circumference to wall thickness ratio that provides an accurate measure of vasospasm and would correct any potential error due to vessel deformity [47]. Interestingly, we found that a 250ppm CO treatment significantly improved vasospasm in mice by increasing the lumen area to wall thickness ratio and lumen circumference to wall thickness ratio in the MCA region. Although there was no significant difference in the lumen area to wall thickness ratio in the ACA region, there was a significant difference in lumen circumference to wall thickness ratio in the MCA region. Apart from calculating the ratio to confirm the significant difference in lumen area and lumen circumference in SAH+Air and SAH+CO-treated groups, we also calculate the value of lumen area, lumen circumference, and wall thickness individually and found that there was significant increase lumen area and lumen circumference and decrease in wall thickness in the MCA region of CO-treated mice as compared to the Air-treated SAH mice. However, there was no significant difference observed in lumen circumference between Air-treated SAH and CO-treated SAH mice although there was significant reduction in wall thickness and lumen area in the ACA region in CO-treated SAH mice. These data indicate the potential direct or indirect vasoprotective mechanism of CO on SAH-induced vasospasm. Further work is ongoing to pinpoint the cascade of events leading to such outcomes.

In terms of mechanisms, CO has emerged to have many functions including vasodilation, neurotransmission, inhibition of platelet aggregation, anti-proliferative effects on smooth muscle, as well as being an anti-inflammatory agent under certain conditions and at various concentrations. In addition of CO being a vasodilator, it can have direct and indirect cytoprotective properties. Furthermore, it is possible that CO could counteract some on the intrinsic properties of SAH blood by-product. For example, various blood products have pro-inflammatory action and CO has been postulated to have anti-inflammatory properties notably by regulation of various kinases. Also, such blood products can trigger oxidative stress, and CO has been shown to regulate various enzymatic pathways that can regulate oxidative stress damage. CO is also likely a play in the reorganization of the changes recorded in the blood vessels anatomy. Furthermore, the oxidized forms of bilirubin may have vasoconstrictive actions that could be mitigated by CO. And, while there is no consensus on the key molecules and pathways involved in SAH-dependent vasospasm, CO can likely also interfere in that process to reduce the impact of vasospasm, and early and delayed brain injury. While CO is produced endogenously through the degradation of heme, it is likely that in the context of SAH there is not enough being produced when needed, and that targeted and timely exogenously delivery would be key. Additional effort is warranted to determine the exact mechanisms involved.

In conclusion, using the endoperforation model, we demonstrated that SAH induces vasospasm, neurological deficit, motor function, and ambulatory activity. Interestingly, intermittent treatment with 250ppm CO daily for 1h until day 7 significantly protects mice from SAH induced vasospasm and neurobehavioral deficit. Although these data establish the therapeutic potential of CO-treatment to limit vasospasm, functional deficit, and brain damage; additional studies are needed to unravel the mechanism of CO-mediated neuroprotection. Considering the particular etiopathology of SAH, we think it is the clinical target of choice to test the optimal therapeutic efficacy of CO.

Abbreviations:

ACA

Anterior cerebral artery

CO

Carbon monoxide

CV

Cerebral vasospasm

HO1

Heme oxygenase 1

MCA

Middle cerebral artery

NDS

Neurological deficit score

SAH

Subarachnoid hemorrhage