Sage Journals: Discover world-class research

Abstract

Molecular mechanisms of cerebral vasospasm after subarachnoid hemorrhage (SAH) include specific modes of cell signaling like activation of nuclear factor (NF)-kB and vascular cell adhesion molecules (VCAM)-1 expression. The study's hypothesis is that cisternal cerebral spinal fluid (CSF) from patients after SAH may cause Ca²⁺ oscillations which induce these modes of vascular inflammation in an in vitro model of human cerebral endothelial cells (HCECs). HCECs were incubated with cisternal CSF from 10 SAH patients with confirmed cerebral vasospasm. The CSF was collected on days 5 and 6 after hemorrhage. Cytosolic Ca²⁺ concentrations and cell contraction as an indicator of endothelial barrier function were examined by fura-2 microflurometry. Activation of NF-κB and VCAM-1 expression were measured by immunocytochemistry. Incubation of HCEC with SAH-CSF provoked cytosolic Ca²⁺ oscillations (0.31 ± 0.09 per min), cell contraction, NF-κB activation, and VCAM-1 expression, whereas exposure to native CSF had no significant effect. When endoplasmic reticulum (ER) Ca²⁺-ATPase and ER inositol trisphosphate (IP3)-sensitive Ca²⁺ channels were blocked by thapsigargin or xestospongin, the frequency of the Ca²⁺ oscillations was reduced significantly. In analogy to the reduction of Ca²⁺ oscillation frequency, the blockers impaired HCEC contraction, NF-κB activation, and VCAM-1 expression. Cisternal SAH-CSF induces cytosolic Ca²⁺ oscillations in HCEC that results in cellular constriction, NF-κB activation, and VCAM-1 expression. The Ca²⁺ oscillations depend on the function of ER Ca²⁺-ATPase and IP3-sensitive Ca²⁺ channels.

Keywords

human cerebral endothelical cells subarachnoid hemorrhage inflammatory reaction;NF-κB VCAM-1

Introduction

After rupture of an intracranial aneurysm that causes subarachnoid hemorrhage (SAH), patients' prognosis is determined by a variety of pathophysiologic mechanisms like vascular inflammation, cerebral vasospasm (CVS), and ischemia (Claassen et al, 2002; Dorsch, 1995). Alterations of function in cerebral vascular endothelial cells seem to be involved in the pathophysiology of all these phenomena (Frijns et al, 2002). Endothelial cells are known to modulate immune defense, coagulation, permeability, and the contraction of the vascular smooth muscle cells and thus vascular tone (Faraci and Heistad, 1998; Iuliano et al, 2004). Many of the endothelial functions and intracellular signal transduction pathways are Ca²⁺ dependent (Tran et al, 2000).

Recently, cytosolic Ca²⁺ oscillations were described as a specific mode of intracellular signal transduction. The underlying mechanism was first described in the De-Young and Kaiser model (Jafri and Keizer, 1995) as a periodic Ca²⁺ release from the endoplasmic reticulum (ER) to the cytosol through activation of the inositol trisphosphate (IP3)-gated Ca²⁺ channel and Ca²⁺ reuptake into the ER by Ca²⁺-ATPase. Meanwhile Ca²⁺ oscillations are regarded as a universal mode of cell signaling that increases signal specificity, prevents desensitizing, and amplify signals in case of low-level stimuli (Dolmetsch et al, 1998). Ca²⁺ oscillations occur in many nonexcitable cells like T-cells, keratocytes, and endothelial cells (Kuebler et al, 2002; Paltauf-Doburzynska et al, 2000). Ca²⁺ oscillations can be induced in human umbilical vascular endothelial cells and human aortic endothelial cells by histamine, thrombin, bradykinin, ATP, or hydrogen peroxide (Hu et al, 1998; Paltauf-Doburzynska et al, 2000). The frequency of Ca²⁺ oscillations has been found to be 0.7 ± 0.1 per min in aortic endothelial cells stimulated by 100 μmol/L hydrogen peroxide and 0.3 per min by 1 μmol/L histamine (Hu et al, 1998).

The activation of NF-κB has been linked to Ca²⁺ oscillations (Dolmetsch et al, 1998). Nuclear factor (NF)-kB is a ubiquitous proinflammatory transcription factor that is activated by degradation of the inhibitory subunit IkB. It regulates the endothelial inflammatory response by inducing the gene expression of adhesion molecules like intercellular adhesion molecules, vascular cell adhesion molecules (VCAMs), chemotactic cytokines like interleukins (ILs) IL-1, IL-6, and IL-8 as well as tumor necrosis factor-α (Dolmetsch et al, 1998; Faraci and Heistad, 1998). However, VCAM-1 expression may also be induced by a variety of other mechanisms. (Chen et al, 1995; Keegan and Zamorano, 1998) VCAM-1 is a member of a family of leukocyte adhesion molecules that are expressed on the luminal surface of endothelial cells and are involved in monocyte adhesion to endothelial cells. This may result in the transmigration of T-cells and monocytes into the vessel wall (De Martin et al, 2000). In addition, both NF-κB and VCAM-1 have been shown to be involved in the pathogenesis of CVS after SAH (Kaynar et al, 2004; Polin et al, 1998; Zhou et al, 2007).

The intracellular Ca²⁺ concentration regulates the function of the endothelial barrier between blood and the surrounding tissue. Ca²⁺ is an activator of myosin light chain kinase, whose activation leads to phosphorylation of myosin light chains and subsequently to actin—myosin interaction. This interaction induces contraction of cells and an increased permeability of the endothelial cell layer (Tiruppathi et al, 2002).

After SAH, endothelial cells are exposed to the bloody cerebral spinal fluid (CSF) because of a direct CSF circulation from the subarachnoid space to the subendothelial layer (Ohta et al, 1992). Macromolecules diffuse within a few minutes to the endothelial cells (Uemura et al, 1987; Yamamoto et al, 1991). The effect of SAH-CSF on the function of cerebral endothelial cells is still unknown. To date, only one publication investigated the influence of erythrocyte lysate and bloody CSF on bovine endothelial cells and showed an adenosine-dependent rise of the cytosolic Ca²⁺ concentration (Zhang et al, 1996).

The goal of this study was to test the hypothesis whether cisternal CSF from patients who experience aneurysmal SAH may cause vascular injury that is initiated by intracellular Ca²⁺ oscillations and modulated by the activation of NF-κB and expression of VCAM-1 in an in vitro model of human cerebral endothelial cells (HCECs). In addition, it is hypothesized that Ca²⁺ oscillations lead to a contraction of HCEC and to an impaired barrier function.

Materials and methods

Cerebral Spinal Fluid and Patient Population

Samples of CSF were derived from a total of 10 SAH patients (male to female = 5:5) with a mean age of 58 ± 11 years (range 40 to 73 years). Patients were included when thick blood clots were seen in the basal cisterns, or intraparenchymal or intraventricular hemorrhage (Fischer 3 or 4) was present on their first computerized tomography scan. In addition, patients underwent surgery for clipping. At the end of surgery, a catheter was placed into the basal cerebral cisterns. Cisternal CSF was sampled on day 5 or 6 after hemorrhage. All patients included experienced global CVS which was confirmed by cerebral angiography on posthemorrhage day 9 ± 2.

For control studies, native CSF was sampled from patients who underwent myelography for the diagnosis of degenerative spinal diseases. In the control group the mean age of the patients was 63 ± 6 years (65 to 75 years).

The CSF from both SAH patients and controls was centrifuged (3,000 g for 10mins) and stored at −80°C upon analysis.

Culture of Human Cerebral Endothelial Cell

The HCEC (Cell Systems Corporation, Kirkland, WA, USA) were cultured in Cell Systems Complete Medium. This medium served for all experiments. The culture medium was changed every 48 h.

Ca2+ Measurement

The cytosolic concentration of Ca²⁺ was measured with the fluorescent dye fura-2 AM. Cells were incubated with 2.5 μmol/L of the acetoxymethyl ester of fura-2 for 60 mins. Fura-2 fluorescence was analyzed using a TILL Photonics imaging system (Gröfelfing, Germany). Illumination was induced by exposing the cells to an excitation light with a wavelength that alternated between 340 and 380 nm. Emitted light was detected at 510 nm. Fura-2 fluorescence was calibrated according to the methods described by Grynkiewicz et al (1985). Cells were exposed to 5 μmol/L ionomycin in modified 4-(2-hydroxyethyl)-1-piperazi-neethanesulfonic acid (HEPES) buffer containing either 3 μmol/L Ca²⁺ or buffer that was Ca²⁺ free with 500 μmol/L ethylene glycol tetraacetic acid to obtain the maximum (R_max) and the minimum (R_min) of the ratio of fluorescence (R), respectively. Intracellular Ca²⁺ concentration was calculated according to the following equation:

[{Ca}^{2 +}]_{i} = K_{d} \times β \times (R - Rmin) / (Rmax - R)

where K_d is the dissociation constant of fura-2 determined in intact cells and ᵝ is the ratio of the 380 nm excitation signals of ionomycin-treated cells at 5 mmol/L ethylene glycol tetraacetic acid and at 3mmol/L Ca²⁺.

Blockers of Endoplasmic Reticulum Ca2+-ATPase and Ca2+ Channels

In the present investigation, ER Ca²⁺-ATPase, IP3-sensitive Ca²⁺ channels, and ryanodine-dependent Ca²⁺ channels were selectively blocked by 50 μmol/L thapsigargin, 10 μmol/L xestospongin, and 10 μmol/L ryanodine, respectively. The optimal concentrations of the blockers were determined from dose—response curves (data not shown). Because xestospongin and thapsigargin are known to cause apoptosis and necrosis, it was tested whether the concentrations of ER blockers used may influence cell viability. Human cerebral endothelial cells were exposed to 50 nmol/L thapsigargin, 10 μmol/L xestospongin, 10 μmol/L ryanodine, and HEPES-buffered saline solution. Apoptosis and necrosis were evaluated by staining with Hoechst-3342 and propidium iodide as previously described (Kumar et al, 2007). For quantitative assay, a blind analysis of 180 to 250 nuclei from randomized 5 fields of view was applied.

Cellular Constriction

Cellular constriction was quantified by planimetrical analysis of the Ca²⁺ independent fluorescence intensity at 360 nm in parallel with the measurement of the fura-2 ratio. Before the addition of CSF (at baseline), the absolute value of cell-surface area was set at 100%. After the addition of CSF the cell-surface area was measured every 6 secs and expressed as a percentage of baseline.

Detection of Nuclear Factor-kB Activity and Vascular Cell Adhesion Molecules-1 Expression

Immunocytochemistry was used to measure activated NF-κB and VCAM-1 expression similar to the method described by Abdallah et al (2007). Human cerebral endothelial cell were stimulated with SAH-CSF in the absence or presence of xestospongin or thapsigargin. Control experiments were performed with native CSF or HEPES-buffered saline solution. After 60 mins stimulation, cells were fixed and permeabilized with ice-cold methanol, followed phosphate-buffered saline/bovine serum albumin blocking. Subsequently, the endothelial cells were incubated with antibodies increased against VCAM-1 or phospho-specific human pS 276 p65 subunit of the activated NF-κB (Ashikawa et al, 2002). The antibodies were obtained from Rockland Laboratories (Gilbertsville, PA, USA). After 60 mins, the endothelial monolayers were washed with phosphate-buffered saline before being incubated with a secondary Cy 2-conjugated goat anti-rabbit immunoglobulin G (1:200) overnight. On the following day, the cells were washed and mounted on coverslips. The fluorescence signal was analyzed by fluorescence microscopy (TILL Photonics) using TillvisION software. All steps were performed at room temperature.

Data Analysis and Statistics

Data are reported as mean ± s.d. or percentage of control. Comparisons among different treatment groups were analyzed with analysis of variance. A value of P < 0.05 was considered statistically significant. The effects of the blockers were tested by paired Students' t-test.

Results

Effect of Subarachnoid Hemorrhage-Cerebral Spinal Fluid on the Cytosolic Endothelial Ca2+ Concentration

Figure 1 shows the time course of the fura-2 ratio in a HCEC monolayer. The brightness of the micrographs is proportional to the Ca²⁺ concentration. The HCECs were incubated in modified HEPES buffer solution for 10 mins. Then the solution was replaced by CSF of SAH patients. Cytosolic Ca²⁺ oscillations appeared with a latency period of 62 ± 25 secs in all the 10 patients. The mean frequency was 0.31 ± 0.09 per min and the oscillations persisted during the whole experiment. Within the oscillatory peaks the cytosolic Ca²⁺ concentrations were 4.8-fold higher compared with the basal concentration (327 ± 98 versus 68 ± 12 nmol/L) (P < 0.001, n = 10). There was variability in the intracellular Ca²⁺ concentration both over time and across cells (Figure 1A). When the endothelial cells were incubated with native CSF, no Ca²⁺ oscillations occurred (Figure 1B). This finding was consistently reproduced in eight experiments.

Figure 1

Ca²⁺ oscillation in fura-ratio pictures and Fura-2 Ratio recordings after exposing cells to CSF. Photomicrographs with fura-2 staining are shown at three different time points after starting the measurement. The brightness of pixels is directly proportional to the Ca²⁺ concentration within the pixel. CSF was added 10 mins after starting the experiment at 0 min. During the observation period, the cytosolic Ca²⁺ concentration is highly variable within and across cells. The maximum cytosolic Ca²⁺ concentration was reached periodically approximately every 3.2 mins, resulting in an oscillation frequency of 0.31 ± 0.02 per min. Each graph showing the fura-2 ratio (in arbitrary units, a.u.) represents the Ca²⁺ concentration of one recorded single cell. After an initial incubation for 10 mins the bath solution was replaced by CSF. (A) Human cerebral endothelial cells were incubated in SAH-CSF from a patient with cerebral vasospasm and (B) HCECs were incubated in native CSF.

The Origin of Subarachnoid Hemorrhage-Cerebral Spinal Fluid-Induced Cytosolic Ca2+ Oscillations

Under different physiologic conditions Ca²⁺ oscillations in endothelial cells originate from disturbances of ER Ca²⁺-channel function and Ca²⁺-ATPase activity. Ca²⁺-ATPase and IP3- or ryanodine-dependent Ca²⁺ channels are the primary candidates for the origin of Ca²⁺ oscillations. When thapsigargin was added to the SAH-CSF 20 mins after the onset of cytosolic Ca²⁺ oscillations, Ca²⁺ oscillations ceased immediately and almost completely. This was consistently reproduced in all five studies. Figure 2 shows the effect in a representative cell.

Figure 2

Effects of ryanodine, xestospongin and thapsigargin on Ca²⁺ oscillations. Ryanodine-dependent Ca²⁺ channels, IP3-dependent Ca²⁺ channels, and Ca²⁺-ATPase were blocked by ryanodine, xestospongin, and thapsigargin, respectively. All three inhibitors were added 20 mins after the onset of Ca²⁺ oscillations. (A—C) Each graph represents the cytosolic Ca²⁺ concentration recorded on a representative single endothelial cell. After the addition of thapsigargin to the SAH-CSF, Ca²⁺ oscillations ceased completely, whereas xestospongin reduced the frequency of Ca²⁺ oscillations significantly to 54% ± 18% of control. Ryanodine had no effect on the frequency of Ca²⁺ oscillations (A).

The addition of xestospongin to the SAH-CSF reduced the oscillation frequency significantly to 54% ± 18% (P < 0.001) of the control frequency, whereas the addition of ryanodine did not lead to a significant change in Ca²⁺ oscillation frequency (104%± 24%).

These results are not because of the toxicity of the blockers and consecutive cell death. In control experiments, where HCECs were exposed to blockers without SAH-CSF for 2 h, the rate of necrosis was 0.3%, 0.5%, and 0.3% for thapsigargin, xestospongin, and ryanodine, respectively. The rate of apoptosis was 0.1%, 0.12%, and 0.33% for thapsigargin, xestospongin, and ryanodine, respectively. In HCECs that were incubated in HEPES-buffered saline solution the rate of necrosis and apoptosis was 0.3% and 0.2%, respectively. The differences between groups were not statistically significant.

Cellular Constriction

When endothelial cells were exposed to SAH-CSF as shown in Figure 3, there was a rapid decrease in cell-surface area because of the shrinking of cells. Twenty minutes after the onset of the oscillations, the cell-surface area was reduced to 79% ± 5% of the area before the addition of SAH-CSF. Sixty minutes after the onset, the reduction was 71%± 6%. Native CSF did not result in a significant change in the surface area. When 10 μmol/L xestospongin or 50 nmol/L thapsigargin were added 20 mins after the stimulation of the cells by SAH-CSF, the cells expanded again. Forty minutes after addition of the blockers, the cell-surface area recovered to 87% ± 4% and 98%± 4% of control, respectively (Figure 3).

Figure 3

Gap formation. After replacement of the HEPES-bufferd saline solution by CSF, the cell-surface area shrinks to 71% ± 6% of control within 60 mins (d). Blocking the Ca²⁺ oscillations by xestospongin (c) and thapsigargin (b) led to a reexpansion of the cells. Forty minutes after the addition of the blockers, the cell-surface area reached 87% ± 4% and 98% ± 4% of control, respectively. When the cells were incubated in native CSF (a), the cell-surface area was nearly constant over the time.

Nuclear Factor-kB and Vascular Cell Adhesion Molecules-1

When activation of NF-κB in HCEC was measured after incubation with HEPES-buffered saline solution and native CSF, there was no significant difference between results with these substances. However, when HCECs were exposed to SAH-CSF, NF-κB activation (Figure 4) increased to 205% ± 28% of control (P < 0.001). This phenomenon could be dampened after preincubation of the cells with 10 μmol/L xestospongin or 50 nmol/L thapsigargin, which reduced the SAH-CSF-induced NF-κB activation to 163% ± 32% and 138% ± 28% of the basal activity, respectively. These results were statistically significant, as shown in Figure 5. After the same protocol, SAH-CSF induced VCAM-1 expression (Figure 4) to 181% ± 19% of basal fluorescence, whereas in the presence of xestospongin and thapsigargin VCAM-1 expression reached only 119% ± 24% and 129% ± 19% of control, respectively. All differences were statistically significant (P < 0.001; Figure 5).

Figure 4

Immunocytochemistry. Human cerebral endothelial cells are labeled by antibodies against VCAM-1 or the phospho-specific human pS 276 p65 subunit of activated NF-κB. The cells were exposed to SAH-CSF, which caused VCAM-1 expression (B), activation of NF-κB and its translocation into the cellular nucleus (A). For control experiments native CSF was used (C and D).

Figure 5

NF-κB activity and VCAM-1 expression. Immunofluorescence was used to detect activated NF-κB and expressed VCAM-1 in human cerebral endothelial cells after being incubated with HEPES-buffered saline solution and CSF, both native and after SAH. Note that CSF from SAH patients nearly doubled the activation of NF-κB. The addition of xestopongin and thapsigargin to CSF from SAH patients reduced the activation of NF-κB and VCAM-1 expression significantly in a close analogy to the reduction of Ca²⁺ oscillations.

Discussion

This study investigates the signaling cascade leading from SAH to a rudimentary endothelial inflammatory reaction after SAH (Figure 6). The principal findings are summarized in the following steps:

(1)

Subarachnoid hemorrhage-cerebral spinal fluid sampled from patients with confirmed CVS induces cytosolic Ca²⁺ oscillations in HCEC.

(2)

Ca²⁺ oscillations originate from the ER by repetitive uptake and release of Ca²⁺ by the ER Ca²⁺-ATPase and IP3-dependent Ca²⁺ channels.

(3)

Ca²⁺ oscillations cause reversible shrinking of HCEC, that is, impaired barrier function.

(4)

The Ca²⁺ oscillations activate proinflammatory NF-κB and lead to increased VCAM1 expression.

Figure 6

Scheme of SAH-induced vascular injury. In HCEC, SAH-CSF-induced cytosolic Ca²⁺ oscillations lead to two distinct pathways that both contribute to vascular injury. On one side impaired endothelial barrier function occurs as a result of cellular shrinking. On the other side activation of NF-κB and expression of VCAM-1 may be regarded as a rudimentary inflammatory reaction.

Mechanisms of Intracellular Ca2+ Homeostasis and Signal Transduction

The ER and mitochondria serves as intracellular Ca²⁺ stores. The main store is the ER, containing 75% of the intracellular Ca²⁺. The mitochondria contain the remaining 25% (Tran et al, 2000). There are a variety of mechanisms that influence intracellular Ca²⁺ homeostasis. Ca²⁺-ATPase primarily acts by pumping Ca²⁺ continuously into the ER lumen. Endoplasmic reticulum Ca²⁺ release is IP3 dependent (Jafri and Keizer, 1995). Inositol trisphosphate binds to specific receptors on the ER surface and activates IP3-gated Ca²⁺ release channels, which results in an increase in the cytosolic Ca²⁺ concentration. The IP3-gated Ca²⁺ channels are the predominant ER Ca²⁺ channels in nonexcitable cells. The ER ryanodine-dependent Ca²⁺ releasing channel is of little importance in endothelial cells. These channels have been most extensively characterized in muscle cells (Tran et al, 2000). In many cell types the IP3-gated Ca²⁺ channels are in a closed loop with the Ca²⁺-ATPase that is responsible for cytosolic Ca²⁺ oscillations.

The Role of Ca2+ Oscillations in Normal Physiology and after Subarachnoid Hemorrhage

Ca²⁺ oscillations trigger universal processes such as cell division, cell differentiation, and apoptosis (Miyakawa-Naito et al, 2003). The frequency or the amplitude or both of Ca²⁺ oscillation control the specificity of this signal (Dolmetsch et al, 1998).

Cytosolic Ca²⁺ osillations are observed in different phenotypes of endothelial cells. In aortic endothelial cells, Hu et al, 1998 reported a frequency of 0.7 ± 0.1 per min stimulated by 100 μmol/L hydrogen peroxide and 0.3 per min by 1 μmol/L histamine. As shown in the present paper, Ca²⁺ oscillations in HCEC occur at an average frequency of 0.31 ± 0.02 per min after exposure to CSF from SAH patients. This frequency is higher than 0.1 per min, which has been described as the cutoff frequency for the exponential activation of NF-κB (Dolmetsch et al, 1998).

Inositol trisphosphate-dependent Ca²⁺ oscillations propagate from one endothelial cell to another (Boitano et al, 1992). Vascular endothelial cells interfere with the surrounding smooth muscle cells through gap junctions. Thus, an increased endothelial cytosolic Ca²⁺ concentration may result in vessel constriction. This has been shown in an in vitro hamster model where a strong vasoconstriction was induced by an increase in endothelial Ca²⁺ concentration (Yashiro and Duling, 2000). The propagation of the Ca²⁺ signal was stopped by thapsigargin (Yashiro and Duling, 2003). The mechanisms involved in regulation of cytosolic calcium are dramatically different in endothelial and vascular smooth muscle cells, and effects in one cell type cannot be extrapolated to the other. It remains speculative whether Ca²⁺ oscillations are involved in the development of CVS.

The specific factors in SAH-CSF that caused Ca²⁺ oscillations in the present investigation are not entirely clear. Previous investigations suggested that oxygen radicals induce endothelial cytosolic Ca²⁺ oscillations by inhibiting the normal function of ER Ca²⁺ channels and pumps (Hu et al, 1998). Oxygen radicals directly attack the Ca²⁺-ATPase at the ATP-binding site (Xu et al, 1997) and inactivate the Ca²⁺ pump. In contrast, others hypothesize that free oxygen radicals activate phospholipase C. Phospholipase C phosphorylates inositoldiphosphate to IP3 and activates the IP₃-dependent Ca²⁺ channels at the ER (Berridge, 1993). The sources of oxygen free radicals after SAH may be the perivascular oxyhemoglobine and the endothelial nicotinamide adenine dinucleotide phosphate-oxidase (Kim et al, 2002; Paravicini and Sobey, 2003). In addition, thrombin and ATP may also induce Ca²⁺ oscillations both alone and in concert. Both are proven to increase levels of IP3, which is linked to ER IP3-dependent Ca²⁺-channel activity (Glass and Bates, 2004; Tiruppathi et al, 2002). After SAH, CSF contains both substances (Tsurutani et al, 2003; Zhang et al, 1996).

The frequency of SAH-CSF-induced Ca²⁺ oscillations are in the range of frequencies that have been shown to activate NF-κB (Dolmetsch et al, 1998). In our experiments the frequency of oscillations was reduced with xestospongin and thapsigargin to 54%± 17% and 0% of control, respectively. In addition, a significant reduction of NF-κB activation was observed that was similar to the reduction of the frequencies of Ca²⁺ oscillations after xestospongin or thapsigargin addition. According to these results, it may be stated that SAH-CSF activates NF-κB, depending on the frequency of Ca²⁺ oscillations. These findings confirm the results of Hu et al who used xestospongin to reduce the frequency of Ca²⁺ oscillations when oscillations were induced by free oxygen radicals. As a result, a significant decrease of NF-κB activity was observed.

The activation of NF-κB is necessary for the induction of an inflammatory reaction. Activation of this factor leads to gene expression of intercellular adhesion molecule-1, VCAM-1, tumor necrosis factor-α, IL-1, IL-6, and IL-8. Tumor necrosis factor-α causes ET-1 gene expression (Dolmetsch et al, 1998; Faraci and Heistad, 1998). The expression of adhesion molecules and chemotactic cytokines can cause T-cells and monocytes to transmigrate into the vessel wall (De Martin et al, 2000). The perivascular-activated monocytes after SAH have been reported as one major source of ET-1 that can cause delayed CVS (Fassbender et al, 2000). There may be a possible link of CVS to SAH-induced Ca²⁺ oscillations through inflammation, because inactivation of NF-κB by pyrolidine dithiocarbamate or antisense oligonucleotids attenuates vasospasm in animal models of SAH (Ono et al, 1998; Zhou et al, 2007). In analogy to the reduction of Ca²⁺ oscillation frequency and NF-κB activation, xestospongin, and thapsigargin reduced the expression of VCAM-1. Ca²⁺ oscillations seem to be important in inducing vascular inflammation after SAH.

The Endothelial Barrier Function

The endothelial contractile mechanism is regulated through phosphorylation of the endothelial cell myosin light chain, which is controlled by Ca²⁺/ calmodulin-activated myosin light chain kinase. The cellular shrinking is Ca²⁺ dependent and subsequently leads to increased vessel permeability. The vessel permeability is proportional to the size of the intercellular space that opens up after cellular shrinking (Schnittler et al, 1990). According to our experiments, Ca²⁺ oscillations were responsible for the development of HCEC constriction. In control experiments with native CSF, no significant change in the intercellular space was observed. Human cerebral endothelial cell constricted when SAH-CSF-induced Ca²⁺ oscillations were present. In parallel with the blockade of the SAH-CSF-induced Ca²⁺ oscillations by thapsigargin and xestospongin, the cells reexpanded after the addition of blockers. These results are consistent with previous findings suggesting that plasmin and thrombin induce cellular constriction in cerebral vascular endothelial cells. The investigators assumed a Ca²⁺-dependent mechanism (Nagy et al, 1995).

After endothelial shrinking, vasoactive substances may diffuse directly from the lumen into the vessels' smooth muscle layer and cause vasoconstriction. This has been shown by Iuliano et al, 2004 in a primate animal model.

The increase of permeability in the blood-brain barrier has also been linked to the formation of brain edema in SAH models of both animals and humans (Xi et al, 2001). Together with VCAM-1, cellular constriction may facilitate the transmigration of monocytes into the vessel wall where these cells can generate one of the most potent vasoconstrictors that are endothelin-1 and others (Fassbender et al, 2000).

Conclusions

After SAH, CSF induces a vascular inflammatory response through Ca²⁺ oscillations in human cerebral endothelial cells. The oscillations depend on ER IP3-receptor-gated Ca²⁺ channels and ER Ca²⁺-ATPase activity. They lead to cellular constriction, NF-κB activation, and VCAM-1 expression.

References

Abdallah

Gligorievski

Kasseckert

Dieterich

Schafer

Kuhlmann

Noll

Sauer

Piper

Schafer

(2007) The role of poly(ADP-ribose) polymerase (PARP) in the autonomous proliferative response of endothelial cells to hypoxia. Cardiovasc Res 73:568–74.

Ashikawa

Majumdar

Banerjee

Bharti

Shishodia

Aggarwal

(2002) Piceatannol inhibits TNF-induced NF-kappaB activation and NF-kappaB-mediated gene expression through suppression of IkappaBalpha kinase and p65 phosphorylation. J Immunol 169:6490–7.

Berridge

(1993) Inositol trisphosphate and calcium signalling. Nature 361:315–25.

Boitano

Dirksen

Sanderson

(1992) Intercellular propagation of calcium waves mediated by inositol trisphosphate. Science 258:292–5.

Chen

Rosenbloom

Anderson

Manning

(1995) Selective inhibition of E-selectin, vascular cell adhesion molecule-1, and intercellular adhesion molecule-1 expression by inhibitors of I kappa B-alpha phosphorylation. J Immunol 155:3538–45.

Claassen

Carhuapoma

Kreiter

Connolly

Mayer

(2002) Global cerebral edema after subarachnoid hemorrhage: frequency, predictors, and impact on outcome. Stroke 33:1225–32.

De Martin

Hoeth

Hofer-Warbinek

Schmid

(2000) The transcription factor NF-kappa B and the regulation of vascular cell function. Arterioscler Thromb Vasc Biol 20:E83–8.

Dolmetsch

Lewis

(1998) Calcium oscillations increase the efficiency and specificity of gene expression. Nature 392:933–6.

Dorsch

(1995) Cerebral arterial spasm—a clinical review. Br J Neurosurg 9:403–12.

10.

Faraci

Heistad

(1998) Regulation of the cerebral circulation: role of endothelium and potassium channels. Physiol Rev 78:53–97.

11.

Fassbender

Hodapp

Rossol

Bertsch

Schmeck

Schutt

Fritzinger

Horn

Vajkoczy

Wendel-Wellner

Ragoschke

Kuehl

Brunner

Schurer

Schmiedeck

Hennerici

(2000) Endothelin-1 in subarachnoid hemorrhage: An acute-phase reactant produced by cerebrospinal fluid leukocytes. Stroke 31:2971–5.

12.

Frijns

Rinkel

Castigliego

Van Gijn

Sixma

Fijnheer

(2002) Endothelial cell activation after subarachnoid hemorrhage. Neurosurgery 50:1223–9; discussion 1229–30.

13.

Glass

Bates

(2004) The role of endothelial cell Ca2+ store release in the regulation of microvascular permeability in vivo. Exp Physiol 89:343–51.

14.

Grynkiewicz

Poenie

Tsien

(1985) A new generation of Ca2+ indicators with greatly improved fluorescence properties. J Biol Chem 260:3440–50.

15.

Corda

Zweier

Capogrossi

Ziegelstein

(1998) Hydrogen peroxide induces intracellular calcium oscillations in human aortic endothelial cells. Circulation 97:268–75.

16.

Iuliano

Pluta

Jung

Oldfield

(2004) Endothelial dysfunction in a primate model of cerebral vasospasm. J Neurosurg 100:287–94.

17.

Jafri

Keizer

(1995) On the roles of Ca2+ diffusion, Ca2+ buffers, and the endoplasmic reticulum in IP3-induced Ca2+ waves. Biophys J 69:2139–53.

18.

Cerebral Blood Flow & Metabolism (2009) 29, 57–65.

19.

Kaynar

Tanriverdi

Kafadar

Kacira

Uzun

Aydin

Gumustas

Dirican

Kuday

(2004) Detection of soluble intercellular adhesion molecule-1 and vascular cell adhesion molecule-1 in both cerebrospinal fluid and serum of patients after aneurysmal subarachnoid hemorrhage. J Neurosurg 101:1030–6.

20.

Keegan

Zamorano

(1998) Regulation of gene expression, growth, and cell survival by IL-4: contribution of multiple signaling pathways. Cell Res 8:1–13.

21.

Kim

Suh

Lee

Kim

Lee

Hong

Kim

(2002) Vascular NAD(P)H oxidase triggers delayed cerebral vasospasm after subarachnoid hemorrhage in rats. Stroke 33:2687–91.

22.

Kuebler

Ying

Bhattacharya

(2002) Pressure-induced endothelial Ca(2+) oscillations in lung capillaries. Am J Physiol Lung Cell Mol Physiol 282:L917–23.

23.

Kumar

Kasseckert

Kostin

Abdallah

Schafer

Kaminski

Reusch

Piper

Steinhoff

Ladilov

(2007) Ischemic acidosis causes apoptosis in coronary endothelial cells through activation of caspase-12. Cardiovasc Res 73:172–80.

24.

Miyakawa-Naito

Uhlen

Lal

Aizman

Mikoshiba

Brismar

Zelenin

Aperia

(2003) Cell signaling microdomain with Na,K-ATPase and inositol 1,4, 5-trisphosphate receptor generates calcium oscillations. J Biol Chem 278:50355–61.

25.

Nagy

Kolev

Csonka

Pek

Machovich

(1995) Contraction of human brain endothelial cells induced by thrombogenic and fibrinolytic factors. An in vitro cell culture model. Stroke 26:265–70.

26.

Ohta

Satoh

Kuroiwa

(1992) The permeability change of major cerebral arteries in experimental vasospasm. Neurosurgery 30:33–15; discussion 335–6.

27.

Ono

Date

Onoda

Shiota

Ohmoto

Ninomiya

Asari

Morishita

(1998) Decoy administration of NF-kappaB into the subarachnoid space for cerebral angiopathy. Hum Gene Ther 9:1003–11.

28.

Paltauf-Doburzynska

Frieden

Spitaler

Graier

(2000) Histamine-induced Ca2+ oscillations in a human endothelial cell line depend on transmembrane ion flux, ryanodine receptors and endoplasmic reticulum Ca2+-ATPase. J Physiol 524(Part 3):701–13.

29.

Paravicini

Sobey

(2003) Cerebral vascular effects of reactive oxygen species: recent evidence for a role of NADPH-oxidase. Clin Exp Pharmacol Physiol 30: 855–859.

30.

Polin

Bavbek

Shaffrey

Billups

Bogaev

Kassell

Lee

(1998) Detection of soluble E-selectin, ICAM-1, VCAM-1, and L-selectin in the cerebrospinal fluid of patients after subarachnoid hemorrhage. J Neurosurg 89:559–67.

31.

Schnittler

Wilke

Gress

Suttorp

Drenckhahn

(1990) Role of actin and myosin in the control of paracellular permeability in pig, rat and human vascular endothelium. J Physiol 431:379–401.

32.

Tiruppathi

Minshall

Paria

Vogel

Malik

(2002) Role of Ca2+ signaling in the regulation of endothelial permeability. Vascul Pharmacol 39: 173–185.

33.

Tran

Ohashi

Watanabe

(2000) Calcium signalling in endothelial cells. Cardiovasc Res 48:13–22.

34.

Tsurutani

Ohkuma

Suzuki

(2003) Effects of thrombin inhibitor on thrombin-related signal transduction and cerebral vasospasm in the rabbit subarachnoid hemorrhage model. Stroke 34:1497–500.

35.

Uemura

Okamoto

Handa

(1987) Disturbance in the intramural circulation of the major Cerebro-pial arteries after experimental subarachnoid haemorrhage. Acta Neurochir (Wien) 89:71–6.

36.

Hua

Bhasin

Ennis

Keep

Hoff

(2001) Mechanisms of edema formation after intracerebral hemorrhage: effects of extravasated red blood cells on blood flow and blood-brain barrier integrity. Stroke 32:2932–8.

37.

Zweier

Becker

(1997) Oxygen-free radicals directly attack the ATP binding site of the cardiac Na+, K(+)-ATPase. Ann N Y Acad Sci 834:680–3.

38.

Yamamoto

Clower

Haining

Asari

Smith

(1991) Adventitial red blood cells produce intimal platelet accumulation in cerebral arteries of cats following subarachnoid hemorrhage. Stroke 22: 373–7.

39.

Yashiro

Duling

(2000) Integrated Ca(2+) signaling between smooth muscle and endothelium of resistance vessels. Circ Res 87:1048–54.

40.

Yashiro

Duling

(2003) Participation of intracellular Ca2+ stores in arteriolar conducted responses. Am J Physiol Heart Circ Physiol 285:H65–73.

41.

Zhang

Weir

Macdonald

Marton

Solenski

Kwan

Lee

(1996) Mechanisms of [Ca++]i elevation induced by erythrocyte components in endothelial cells. J Pharmacol Exp Ther 277:1501–1509.

42.

Zhou

Shi

Hang

Cheng

Mao

Chen

Yin

(2007) Potential contribution of nuclear factor-kappaB to cerebral vasospasm after experimental subarachnoid hemorrhage in rabbits. J Cereb Blood Flow Metab 27:1583–92.

Cytosolic Ca 2+ Oscillations in Human Cerebrovascular Endothelial Cells after Subarachnoid Hemorrhage

Abstract

Keywords

Introduction

Materials and methods

Cerebral Spinal Fluid and Patient Population

Culture of Human Cerebral Endothelial Cell

Ca2+ Measurement

Blockers of Endoplasmic Reticulum Ca2+-ATPase and Ca2+ Channels

Cellular Constriction

Detection of Nuclear Factor-kB Activity and Vascular Cell Adhesion Molecules-1 Expression

Data Analysis and Statistics

Results

Effect of Subarachnoid Hemorrhage-Cerebral Spinal Fluid on the Cytosolic Endothelial Ca2+ Concentration

The Origin of Subarachnoid Hemorrhage-Cerebral Spinal Fluid-Induced Cytosolic Ca2+ Oscillations

Cellular Constriction

Nuclear Factor-kB and Vascular Cell Adhesion Molecules-1

Discussion

Mechanisms of Intracellular Ca2+ Homeostasis and Signal Transduction

The Role of Ca2+ Oscillations in Normal Physiology and after Subarachnoid Hemorrhage

The Endothelial Barrier Function

Conclusions

References