This study is the first to investigate the timecourse of the inflammatory response, specifically changes in the expression of IL-6, IL-1β, TNFα and MMP-9, in the walls of cerebral arteries over the first 4 days following experimental SAH. Furthermore, it is the first to investigate the role of MEK-ERK1/2 activation during the first 24 h post-SAH as an early trigger of the more delayed inflammatory response in the cerebral vasculature.
The study reveals an early transient peak at 6 h post-SAH in IL-6, IL-1β and MMP-9 protein expression in cerebral artery walls followed by a stronger and more persistent increase in cerebrovascular protein expression levels of TNFα, IL-6, IL-1β and MMP-9 at 24 h post-SAH and onwards until 4 days post-SAH where a beginning decline in these markers was observed. Analysis of ERK1/2 phosphorylation in cerebral arteries revealed that at 1 h post-SAH it was already strongly activated, which is in accordance with an earlier western blot study
. The SAH-induced ERK1/2 activation persisted until 3 days post-SAH and thereafter declined at 4 days post-SAH. Thus, induction of ERK1/2 activity is a very early event in cerebral arteries preceding the elevation of cytokines and MMP-9 commencing somewhat later in the SMCs and peaking at 3 days post-SAH. These events are primarily concentrated to the cerebrovascular SMCs not only in the large circle of Willis arteries but also in parenchymal microvessels. This provides the first direct evidence of an associated vascular mechanism, involving both large cerebral arteries and brain microvessels.
Late cerebral ischemia in man typically occurs 4 to 12 days after SAH and involves several pathophysiological mechanisms, including cerebral vasoconstriction
, endothelial dysfunction
[40, 41], blood-brain barrier breakdown
 as well as a marked inflammatory response
[2, 15]. Analysis of genes involved in the events associated with SAH show a spectrum of functional systems such as those of fibrinolysis, inflammation, vascular reactivity and neuronal repair
. We have earlier demonstrated that the expression of several vasoconstrictor receptors is increased in cerebral arteries after SAH, and that this upregulation occurs via early MEK-ERK1/2 activation, similarly to what we demonstrate in the present study for the upregulation of proinflammatory cytokines and MMP-9. We hypothesize that the ERK1/2-mediated inflammatory response and vasoconstrictor receptor upregulation works in concert
 to disturb the normal cerebrovascular reactivity and thereby contribute to development of delayed cerebral ischemia
A large number of earlier studies have investigated the expression of proinflammatory cytokines in ischemic brain tissue where they contribute to elevated infarct size and more neurological deficits. The proinflammatory cytokines TNFα, IL-6 and IL-1β have been found in the brain infarct region
 and the same cytokines have been suggested to be involved in the development of late cerebral ischemia after SAH
[6, 15]. The pattern of cytokine expression differs depending on stroke type and localization. Comparison of two SAH models revealed that the expression of TNFα, IL-6 and IL-1β were differentially increased in the brain at 2 and 7 days post-SAH
. Thus, following perforation of cerebral arteries the response was small and late (at 7 days), while after blood injection into basal cisterns, an increase of cytokines was found in the brain tissue at 2 days after SAH.
In the present study, we found that at early timepoints (1 to 24 h) following SAH there was a marked expression of TNFα in the brain parenchyma. Interestingly, the other two cytokines studied, IL-6 and IL-1β, did not show any enhanced expression in the brain tissue. The elevated brain TNFα expression was colocalized with GFAP, a marker of glial cells and astrocytes around the vessels and in the brain tissue. This is in accordance with the previous observations of TNFα mRNA and protein expression at an initial peak around 1 to 3 h and a second peak at 12 to 24 h in cortical neurons after focal cerebral ischemia
, and in microglia and astrocytes after transient forebrain ischemia in rats
. Activation of microglia and astrocytes leading to release of cytotoxic substances such as nitric oxide and TNFα have been reported in response to early brain injury after cerebral ischemia
. In addition, increased production of TNFα in brain tissue has been reported following cerebral ischemia
. A recent study reported that blockade of endogenous TNFα may significantly reduce infarct size in rats following permanent and transient focal cerebral ischemia, suggesting involvement of TNFα in ischemic neuronal damage
. We speculate that reactive gliosis early after SAH may give rise to the increased brain TNFα levels observed in the present study, which may in turn play a role in early neuronal damage after SAH.
With regard to the expression of cytokines in the walls of the cerebral vasculature, we demonstrate a significantly increased expression of TNFα in the SMCs of the MCA at 2 to 4 days post-SAH. There was moderately increased TNFα expression at 24 h post-SAH, which increased slightly over time during 48 to 96 h post-SAH. This is in agreement with a previous study that showed that TNFα mRNA was elevated at 24 to 48 h in the walls of cerebral arteries after SAH
. In addition, another group of investigators reported on TNFα expression in the wall of the BA at 2 to 5 days post-SAH in mice
. Interestingly, there was no change in TNFα expression at early timepoints (0 to 12 h post-SAH) in the cerebral artery walls, indicating a shift in the SAH-induced TNFα expression from brain tissue at early timepoints to the cerebral vasculature at delayed timepoints.
For IL-6 and IL-1β, we observed a minor early increase in expression of these cytokines in the walls of cerebral arteries and microvessels at the early timepoint 6 h post-SAH, which further increased over time to reach a maximum peak at 72 h post-SAH. The findings are consistent with other investigations, which have demonstrated that IL-6 and IL-1β mRNAs and proteins were increased in cerebral vessels following transient global ischemia
 and SAH
We did not find any enhanced expression of TNFα, IL-6 and IL-1β proteins in fresh rats, 0 h group or sham operated animals. This indicates that the production and secretion of the studied cytokines correlate with the events occurring after SAH.
Cytokines are pleiotropic low molecular weight proteins with multiple diverse biological activities. The production of cytokines after SAH results in the induction of cyclo-oxygenase 2 (COX2), which is involved in the breakdown of arachnoid acid and the activation of the lipoxygenase pathway. COX2 activation after SAH has been suggested to cause cerebral artery vasoconstriction, activation and infiltration of leukocytes and neutrophils, increased vascular permeability, and increase in reactive oxygen species
. Moreover, there is evidence that TNFα activates SMC NADPH oxidase, which again leads to generation of reactive oxygen species, resulting in cerebrovascular constriction and reduced blood flow
. Accordingly, TNFα, IL-6 and IL-1β have been shown to correlate to the severity of SAH, cerebral vasospasm, development of late cerebral ischemia and secondary brain damage in primate
MMP-9 is a member of the matrix metalloproteinase family of proteinases, which play important roles in remodeling of extracellular matrix components (collagen, laminin and elastin) in the walls of blood vessels
. It has been reported that MMP-9 following cerebral ischemia is able to degrade the endothelial basal lamina and thereby increase the permeability of the BBB
. Increased expression of MMP-9 has been observed in cerebral aneurysm walls in humans
. Previously, we reported that MMP-9 mRNA and protein levels were increased in the SMC of cerebral arteries and microvessels at 24 and 48 h after SAH
 and focal cerebral ischemia
[59, 60]. In the present study we revealed an early slight increase in MMP-9 expression in the walls of cerebral arteries at 6 h and it increased to 72 h. This is in agreement with the demonstration of transcriptional MMP-9 mRNA upregulation in vivo. However, there was very weak expression of MMP-9 in the brain tissue at all timepoints, which is in agreement with previous work
. We therefore speculate that the upregulation of MMP-9 is a response to SAH specific for the cerebral vasculature, and that the upregulated MMP-9 may play a role in the complex vasculopathy after SAH.
Several studies have hypothesized on the involvement of the MEK-ERK1/2 pathway in development of cerebral vasospasm after experimental SAH
[62, 63]. We observed a significant increase in ERK1/2 phosphorylation in the wall of the MCA following SAH. Interestingly, this expression started already at the early timepoint 1 h and remained elevated over time until 4 days post-SAH. Ansar and Edvinsson reported that only the MEK-ERK1/2 pathway was activated at early timepoints after SAH, while p38 or c-Jun N-terminal kinase (JNK) were activated only at 48 h after SAH
. Furthermore, Larsen and coworkers
 reported that the specific MEK1/2 inhibitor U0126 prevented SAH-induced late cerebral ischemia in rats. Blockade of the upstream activator of MEK1/2, with a specific Raf inhibitor (SB386023-b), also prevented the enhanced cytokines and MMP-9 expression after SAH, and the associated delayed reduction in CBF
. We found that U0126 significantly aborted the phosphorylation of ERK1/2, which in turn decreased the enhanced expression of IL-6, IL-1β, TNFα and MMP-9 proteins in cerebral arteries and microvessels. Secondly, we tested if our hypothesis that this is due to an early ‘switch-on’ mechanisms and that treatment for only the first day would be enough. Interestingly, treatment with U0126 showed similar effects irrespectively of whether (A) we administered it at 6, 12, 24, 36 h and terminated the experiment at 48 h after SAH or if (B) we administered it at 6, 12 and 24 h and terminated the experiment at 72 h after SAH. This shows that it is irrelevant to the effect of U0126 whether the animals are left untreated from 24 h to 72 h post-SAH as long as they are properly treated in the critical period at 6 to 24 h post-SAH, thereby aborting the early ERK1/2 activation. Moreover, the fact that this and earlier studies have consistently shown that the MEK-ERK1/2 inhibitors do not have to be administered immediately after the SAH to exert their beneficial effects, but can be delayed until 6 h post-SAH, suggest a clinically feasible therapeutic time window for these treatments.
Another important observation was that the inhibition of the ERK1/2 signaling by U0126 improved animal behavior scores, including locomotor function and coordination, and spontaneous activity. This is in contrast to the recent Clazosentan studies that were positive in reducing vessels diameter but failed to improve neurological outcome
[64, 65]. Thus, inhibition of early MEK-ERK1/2 signaling after SAH provides a novel therapeutic target that can be treated within a clinically relevant time window and has the potential of improving outcome after SAH.