Neurosurg Focus 3 (4):Article 5, 1997
The complement membrane attack complex and the bystander effect
in cerebral vasospasm
Charles C. Park, M.D., Ph.D., Moon L. Shin, M.D., and J. Marc Simard, M.D., Ph.D.
Departments of Neurosurgery and Pathology, University of Maryland School of Medicine, Baltimore, Maryland
Activation of complement results in formation of membrane attack complexes (MACs) that can insert themselves either into cells that initiate complement activation or into nearby ("innocent bystander") cells. The MACs form large-conductance, nonspecific ion channels that can cause lytic or sublytic cell damage. The authors used a highly sensitive patch clamp technique to assess the contribution of the bystander effect to the pathophysiology of cerebral vasospasm. They compared the effect of complement activation by autologous aged versus fresh erythrocytes on the membrane conductance of freshly isolated rat cerebral artery smooth-muscle cells. In the presence of autologous serum, aged, but not fresh,
erythrocytes caused a large increase in membrane conductance, an effect that was prevented by
heat-inactivating the serum. Ethyleneglycol tetraacetic acid in the presence of Mg++ attenuated the effect,
indicating that complement activation was taking place via the classic pathway. The effect was
reproduced by zymosan-activated autologous serum, suggesting that such changes in conductance could result from insertion of MACs secondary to a bystander effect. Both C8- and C9-depleted heterologous sera produced minimal effects that were converted to full effect by addition of the missing complement component. Superoxide dismutase plus catalase did not attenuate the conductance changes produced by autologous serum plus aged erythrocytes. Autologous serum plus aged erythrocyte membrane ghosts that were free of lysate caused a typical increase in conductance. This study demonstrates that complement activation by aged erythrocytes can result in MAC insertion into innocent bystander smooth-muscle cell membranes and that this mechanism, heretofore undescribed, may contribute to development of
vasospasm after subarachnoid hemorrhage.
Key Words * subarachnoid hemorrhage * vasospasm * complement * ion channels * bystander effect * rat
Cerebral vasospasm remains an important cause of morbidity and mortality after subarachnoid hemorrhage (SAH). Despite numerous advances, the molecular mechanism that initiates cerebral
vasospasm remains undetermined. Much of the previous work on cerebral vasospasm has focused on two broad types of mechanisms: one involving putative spasmogens released during erythrocyte lysis that affect a physiologically intact vessel, the other involving an inflammatory response in the vessel wall. Excellent reviews of these topics have been presented.[17]
A number of findings indicate that the complement system is essential to the development of vasospasm. Activated components of complement have been found in plasma and cerebrospinal fluid (CSF) of
patients suffering from SAH.[12,25-27] Levels of membrane attack complexes (MACs) are higher in the CSF of patients with SAH whose course is complicated by ischemia-induced neurological deficits.[45] Complement deposits have been identified in the walls of vasospastic vessels in patients[11] and
experimental animals.[10] Most important, blocking of complement activation by the protease inhibitor FUT-175 has been shown to ameliorate vasospasm both in a rabbit model[47,48] and in humans in a therapeutic trial;[49] recently, it was shown that specific depletion of complement by systemic administration of cobra venom abrogates the development of experimental vasospasm in rabbits.[7] After 2 to 3 days of aging ex vivo, initially protected autologous erythrocytes undergo plasma-induced hemolysis known to depend on complement activation.[29,30] A similar process of erythrocyte aging with subsequent complement-mediated hemolysis is postulated to occur in the subarachnoid space after SAH and may account for the initiation and timing of vasospasm. Complement activation is required for appreciable lysis of aged erythrocytes by autologous serum in vitro,[29] and aged erythrocytes are
required for the development of late vasospasm.[30]
Spasmogens in hemolysate, especially oxyhemoglobin and free radicals, have been postulated to be responsible for delayed cerebral vasospasm,[16,18] although specific roles for oxyhemoglobin or free radicals recently have been questioned.[1,15] We hypothesized that cerebrovascular smooth-muscle cells that are adjacent to erythrocytes undergoing complement-mediated hemolysis might be subject to the bystander effect. This phenomenon, which requires activation of complement, comes about when C5b-C7 complexes generated in response to antigenic cells, are inserted nonspecifically into nearby nonantigenic cells (so called "innocent bystander" cells).[8] Subsequent assembly of C8 and C9
completes the formation of MACs, C5b-C9, which are known to act as large-conductance nonspecific ion channels[36] that can initiate lytic or sublytic cell damage.[19-21,32,33] In smooth-muscle cells, sublytic
concentrations of MACs would act as rogue ion channels, admitting Ca++ and other ions into the cell
and, thereby, initiating vasospasm.
We tested the bystander hypothesis using the aged erythrocyte model of Peterson, et al.,[29] to generate MACs. We used a patch-clamp technique, the most sensitive method available for detecting changes in membrane conductance, to monitor MAC insertions into single cells.[9,13] We studied freshly isolated smooth-muscle cells from the basilar artery because these cells maintain their normal contractile
phenotype[44] and, thus, are more suitable for study than cultured cells, which undergo phenotypic alterations in vitro. Here we present evidence that in this model, smooth-muscle cells undergo MAC insertions via a bystander mechanism, and this causes significant changes in membrane permeability that may contribute to the development of vasospasm.
MATERIALS AND METHODS
Animal care and procedures were performed in accordance with a protocol approved by the Institutional Animal Care and Use Committee of the University of Maryland School of Medicine and complied with standards set by the United States Department of Health and Human Services and the National Institutes of Health.
Preparation of Smooth-Muscle Cells, Erythrocytes, and Autologous Serum
described.[37,42] Erythrocytes were isolated and aged according to the protocol of Peterson, et al.[29] Briefly, erythrocytes for aging were obtained by femoral vein cutdown; the cells were heparinized and washed three times in sterile artificial CSF, which contained (in mmol/L): 129 NaCl, 4.1 KCl, 1.54
NaH2PO4, 24.9 N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid (HEPES), 1.19 MgCl2, 1 CaCl2,
and 5 glucose. Prepared cells were stored at 37šC in a humidified incubator with 5% CO2. Prior to use,
aged erythrocytes were washed three times by centrifuging at 500 G for 5 minutes each in the bath solution used for electrophysiology. The rat from which cells were harvested was killed 3 to 4 days later to obtain fresh erythrocytes, autologous serum, and cerebral smooth-muscle cells. Fresh erythrocytes were washed three times by centrifuging at 500 G for 5 minutes each in the bath solution and diluted to a hematocrit of 50% prior to use. The number of erythrocytes was calculated by a spectrophotometric method from the amount of hemoglobin released after lysis.[2] For rat erythrocytes, the absorbance
measurement of 0.791 at a wavelength of 410 nm was equivalent to 108 erythrocytes/ml (unpublished
observation). Autologous serum was obtained by allowing blood to clot for 1 hour at room temperature and harvesting the supernatant after centrifuging at 1000 G for 10 minutes. Serum was kept on ice until used and was heat inactivated by incubation at 57šC for 1 hour.
Ghost Preparation
To cause lysis for preparation of erythrocyte ghosts, 0.5 ml of packed erythrocytes were suspended in 0.5 ml of distilled water, vortexed for 1 minute, and centrifuged at 1000 G for 10 minutes. The pellet was washed three times with distilled water. For each experiment, we used 0.2 ml of a suspension of 0.1 ml
of the ghost preparation in 0.5 ml of bath solution, corresponding to 2.2 X 1010 ghost erythrocyte
membranes.
Reagents Used in the Study
Zymosan was diluted to 4 mg/ml in saline. It was then boiled for 10 minutes, washed two times in sterile saline, and kept frozen at -20šC until used. Before the experiment, 1 ml of zymosan was thawed,
centrifuged, and resuspended in 1 ml of serum. The mixture was warmed for 2 to 3 minutes at 30šC and added to the recording chamber. The zymosan-activated supernatant was prepared by incubating the zymosan with serum at 30šC for 15 minutes and collecting the supernatant after centrifugation. The protein concentrations of C8- and C9-depleted sera were both 40 mg/ml, and 100 µl of each component was used for the experiments.
Electrophysiological Study
Cells were voltage-clamped using the whole-cell patch-clamp technique.[9,13] Membrane currents were amplified and sampled online by using a microcomputer equipped with appropriate software. Pipettes with a tip resistance of 1 to 3 M(omega) were made from borosilicate glass and heat polished. The
pipette was filled with solution containing (in mmol/L): 145 KCl, 2 MgCl2, 5 ethyleneglycol tetraacetic
acid (EGTA), 12.5 glucose, and 10 HEPES, pH 7.2, with the solution being filtered prior to use. The
standard bath solution contained (in mmol/L): 140 NaCl, 5 KCl, 1 MgCl2, 12.5 glucose, and 10 HEPES,
pH 7.3. The initial recording chamber volume was 0.25 ml and, with the addition of subsequent
components (0.2 ml each of erythrocytes and serum), the final volume was 0.65 ml. Experiments were performed at room temperature. Current records were filtered at 1 kHz.
After placing freshly isolated smooth-muscle cells in the recording chamber, a whole-cell seal was established to one cell and voltage pulses were started. Once the seal had stabilized, usually within 5 minutes, various components were added sequentially to the recording chamber. To isolate current from complement channels, the holding potential was set at 0 mV to inactivate voltage-dependent channels. Voltage steps to -10 mV and then to +10 mV were given every 2 seconds and the resulting currents were amplified and recorded for up to 2 hours. The pulse protocol used for measuring input conductance is shown in Fig. 1A.
Fig. 1. Traces showing voltage pulse protocol and resulting current. A: Voltage pulse with initial holding potential of 0 mV. A voltage step to -10 mV was applied for 320 msec
followed by a step to +10 mV for another 320 msec before returning to 0 mV; pulse duration 1.6 seconds; pulse repetition rate 0.5 Hz. B: Current trace after formation of the whole-cell seal. Capacitative transients are present at each of the changes in voltage, indicating
successful formation of a gigaohm seal. Initially, the differences in current amplitudes produced by the positive- and negative-voltage steps were 20 to 50 pA (1-2.5 nS). C: Current trace from the same cell as that shown in B, 15 minutes after addition of aged erythrocytes and autologous serum. The difference in current amplitudes resulting from the same voltage step was 520 pA (26 nS). ms = millisecond.
Data Analysis
Membrane currents were measured as the average current during the middle two-thirds of individual recordings produced by each voltage step. As shown in Figs. 2A and B and 3A and B, currents at -10, 0, and +10 mV were plotted as a function of time. The input conductance of the cell was measured as the difference in current produced during voltage steps at -10 mV and at +10 mV, divided by the pulse amplitude, 20 mV. Typically, the initial input conductance was less than 0.2 nS, and under control
conditions this tight seal could be maintained up to 2 hours without significant change (see Fig. 2C). The steady-state conductances reported in the bar graphs of Figs. 2C, 3C, and 4 were obtained by averaging values after they had become stable, typically 10 to 15 minutes after the addition of the various
Newman-Keuls' multiple comparison test. All values are reported as the mean ± standard deviation.
Fig. 2. A and B: Traces showing changes in current amplitude with fresh versus aged erythrocytes. The three superimposed traces represent sequential current measurements at +10 mV, 0 mV, and -10 mV, respectively. A: Current amplitudes after addition of fresh erythrocytes (first arrow) and autologous serum (second arrow). B: Current amplitudes after addition of aged erythrocytes (first arrow) and autologous serum (second arrow). C: Bar graph displaying steady-state conductances obtained 10 to 15 minutes after addition of various components, calculated by dividing the differences in the amplitudes produced at -10 mV and at +10 mV by 20 mV. Abbreviations: ARBC = addition of 0.2 ml of aged
erythrocytes only; ARBC&HI SERUM = aged erythrocytes and heat-inactivated autologous serum added sequentially; ARBC&SERUM = aged erythrocytes and autologous serum added sequentially; ARBC&SERUM PLUS EGTA = bath solution containing EGTA and
Mg++ prior to adding aged erythrocytes and autologous serum; CNTL = no manipulation
after formation of whole-cell seal; FRBC = addition of 0.2 ml of fresh erythrocytes only; FRBC&SERUM = fresh erythrocytes and autologous serum added sequentially; SERUM = addition of 0.2 ml of autologous serum only.
Sources of Supplies and Equipment
The complement components C8 and C9, as well as the C8- and C9-depleted human sera, were
purchased from Quidel, San Diego, CA; all other chemicals, including zymosan, were purchased from Sigma Chemical Co., St. Louis, MO. Amplification and sampling of membrane currents were performed using a microcomputer with computing materials (DigiData 1200 and pClamp software) manufactured
by Axon Instruments, Foster City, CA. The Kimax pipettes used in the electrophysiological studies were purchased from Fisher Scientific, Pittsburgh, PA; the filters were obtained from Millipore Corp.,
Medford, MA.
Fig. 3. A and B: Traces showing changes in current amplitude with antioxidants and erythrocyte ghosts. The three superimposed traces represent sequential current
measurements at +10 mV, 0 mV, and -10 mV, respectively. A: Current amplitude following addition of superoxide dismutase and catalase (first arrow) followed by aged erythrocytes (second arrow) and autologous serum (third arrow). B: Current amplitudes following addition of aged erythrocyte ghosts (first arrow) and autologous serum (second arrow). C: Bar graph displaying steady-state conductances obtained 10 to 15 minutes after addition of components, calculated by dividing the differences in the amplitudes produced at -10 mV and at +10 mV by 20 mV. Abbreviations: AO = after addition of aged erythrocytes and autologous serum in the presence of superoxide dismutase and catalase; GHOST = aged erythrocyte ghosts and autologous serum added sequentially.
RESULTS
Increase in Steady-State Conductance by Aged Erythrocytes
Because Peterson, et al.,[29] had shown that only aged erythrocytes cause complement activation, we first tested aged versus fresh erythrocytes for effects on smooth-muscle cell membrane conductance. Figure 1 shows individual current traces recorded from a basilar artery cell before (Fig. 1B) and after (Fig. 1C) addition of 3-day-aged erythrocytes in the presence of autologous serum. The large increase in current in response to the ± 10-mV pulse became evident several minutes after addition of the aged erythrocytes. The same phenomenon is evident in the recording of current amplitude as a function of time (Fig. 2B) in another cell during addition of aged erythrocytes (Fig. 2B, first arrow) and autologous serum (Fig. 2B, second arrow). Approximately 10 minutes after addition of autologous serum there was an abrupt change in the current amplitude, which then stabilized. By contrast, when fresh rather than aged erythrocytes were used, there was no significant change in current during the course of the experiment (Fig. 2A).
Changes in current observed in other cells were quantified as changes in steady-state conductance (see Materials and Methods). The steady-state conductance of basilar artery cells measured after exposure to
aged erythrocytes and autologous serum was 30.8 ± 18.2 nS (12 cells; Fig. 2C), which was significantly different from that with fresh erythrocytes and autologous serum (0.89 ± 0.44 nS; 8 cells; Fig. 2C; p < 0.001). The order in which aged erythrocytes or autologous serum was added did not affect the outcome. Also, autologous serum, fresh erythrocytes, and aged erythrocytes were tested individually and were found to have no effect on steady-state conductance of basilar artery cells. Conductance values were: 0.85 ± 0.24 nS for control (six cells); 0.80 ± 0.19 nS for serum alone (four cells); 0.75 ± 0.65 nS for fresh erythrocytes alone (four cells); and 0.66 ± 0.56 nS for aged erythrocytes alone (five cells), which were not statistically different from each other (Fig. 2C).
We used heat-inactivated autologous serum to test the requirement for active complement for causing the large conductance changes observed with aged erythrocytes. By incubating serum at 57šC for 1 hour, the hemolytic activity of complement proteins is selectively neutralized. When used with aged erythrocytes, the steady-state conductance of heat-inactivated autologous serum was not significantly different from control, but was significantly different from noninactivated autologous serum (1.2 ± 0.81 nS; four cells; p < 0.001). These findings indicate the presence of a heat-labile factor, presumably complement, in the autologous serum that interacts only with aged, but not fresh, erythrocytes to cause changes in
steady-state conductance of the smooth-muscle cells.
Further evidence favoring involvement of complement activation was obtained in experiments using 5
mM EGTA, a Ca++ chelator, in the presence of 1 mM Mg++. In this condition, the classic pathway of
complement activation requiring Ca++ is inhibited, but the Mg++-dependent alternative pathway is not. In
the presence of 5 mM EGTA, aged erythrocytes and autologous serum did not produce a significant conductance change (1.1 ± 0.76; four cells; Fig. 2C), suggesting that the conductance changes were due to complement activation by the classic pathway.
Antioxidant Effect
Other studies have suggested that oxyhemoglobin and free radicals may be important in causing vasospasm and that use of the antioxidants superoxide dismutase and catalase can exert a protective effect.[38] To determine whether the changes in the steady-state conductance that we observed could be attributed to oxygen free radicals, we studied the effect of aged erythrocytes plus autologous serum in the presence of superoxide dismutase and catalase.[38] Figure 3A shows membrane current amplitude in an experiment using a solution of superoxide dismutase (0.2 ml of 5000 U/ml, final concentration 1000 U/ml) and catalase (0.15 ml of 2100 U/ml, final concentration of 300 U/ml) (first arrow). The second and third arrows in this figure represent, respectively, the addition of 0.2 ml each of aged erythrocytes and autologous serum, bringing the final chamber volume to 1 ml. With these additions, a change in
conductance occurred that was similar in magnitude (45.2 ± 10.1 nS; three cells; Fig. 3C) to that of aged erythrocytes plus autologous serum without antioxidants (30.8 ± 18.2 nS; 12 cells; Fig. 2C). These
findings indicate that the presence of antioxidants provided no protection against conductance changes in smooth-muscle cells observed with aged erythrocytes plus autologous serum.
Erythrocyte Ghosts
The ability of aged erythrocytes to cause conductance changes in smooth-muscle cells was further examined by preparing erythrocyte ghosts after lysing erythrocytes and removing the lysate. Figure 3B shows membrane current amplitude as a function of time after addition of 3-day-aged erythrocyte ghosts (first arrow) and autologous serum (second arrow). The steady-state conductance values for the aged erythrocyte ghost preparation (29.8 ± 10.6 nS; three cells; Fig. 3C) were similar to those of aged
erythrocytes plus autologous serum, indicating that the lysate was not required to observe these changes in conductance with aged erythrocytes.
Bystander Effect Induced by Zymosan
To examine more fully the hypothesis that MAC insertions caused steady-state conductance changes, we compared changes in the steady-state conductances produced by other methods of complement
activation. Zymosan is known to activate complement by an antigen-independent alternative pathway.[6] When autologous serum was added to zymosan, thereby activating the complement system, there was a significant change in the conductance (39.8 ± 10.9 nS; three cells; Fig. 4A), as compared with the
conductance observed in heat-inactivated autologous serum (1.4 ± 0.9 nS; three cells; Fig. 4A; p < 0.02). To exclude a direct effect of zymosan, activated serum was obtained by first incubating zymosan with serum and then removing the zymosan by centrifugation. When applied to the smooth-muscle cell within 10 to 15 minutes of activation, this activated serum also produced an increase in conductance (51.3 ± 28.4 nS; three cells; Fig. 4A) that was not statistically different from that found with either aged erythrocytes plus autologous serum or zymosan plus autologous serum. The similarity of the changes suggests that they are mediated by the same mechanism, that is, insertion of MACs.
Fig. 4. A: Bar graph depicting bystander effect with zymosan-activated autologous serum. Steady-state conductances obtained 10 to 15 minutes after addition of various components, calculated by dividing the differences in the amplitudes produced at -10 mV and at +10 mV by 20 mV. Abbreviations: Z.HI = zymosan plus heat-inactivated autologous serum; Z.SUP = only supernatant was added after autologous serum was activated by zymosan; ZYMO = zymosan plus autologous serum. B: Bar graph displaying complement activation using C8-and C9-depleted heterologous sera (C8D C8-and C9D, respectively).
Bystander Effect Induced by Heterologous Serum
The conductance changes due to complement were further examined by using complement-depleted heterologous serum. Complement in heterologous serum would be activated by the smooth-muscle cell under study, with the size of the channel formed being dependent on the availability of terminal C8 and C9 factors. Thus, when these factors are missing, the conductance changes should be smaller. This is shown in Fig. 4B, where C8- and C9-depleted human sera moderately increased the conductance to 5.6 ±
3.1 nS (three cells) and 8.8 ± 1.1 nS (three cells), respectively. However, when the missing complement components were added, we observed larger increases that were statistically significant: to 29.9 ± 2 nS (three cells, p < 0.001) when C8 was added to C8-depleted heterologous serum and to 31.5 ± 3.5 nS (three cells, p < 0.001) when C9 was added to C9-depleted heterologous serum. These findings further support the hypothesis that changes in the conductances of smooth-muscle cells observed with aged erythrocytes plus autologous serum were mediated by MAC insertions.
DISCUSSION
The model system we developed to test the bystander hypothesis originates from the work of Peterson, et al.,[29] who described initiation of complement activation in vitro with aged erythrocytes and autologous serum. We added freshly isolated smooth-muscle cells to this construct to reproduce in vitro the essential cellular elements likely to be involved in vasospasm in vivo. Using patch-clamp methods, this system allows direct measurements of membrane currents in isolated smooth-muscle cells.
We found that addition of aged, but not fresh, erythrocytes in the presence of autologous serum can cause an increase in membrane conductance of smooth-muscle cells. The hypothesis that changes in
steady-state conductance with aged erythrocytes required activation of serum complement is supported by a number of observations, including: 1) a lack of response with heat-inactivated serum; 2) the ability
of EGTA to block the effect of serum, which demonstrates a requirement for Ca++ to activate the classic
complement pathway; and 3) the similarity in the magnitude of the conductance changes with
zymosan-activated serum and with heterologous serum in the absence of erythrocytes. Our results are in accord with those of Peterson, et al.,[29] who reported that only aged erythrocytes activate the
immunological reaction. It is not known why aged erythrocytes cause autologous complement activation, but expression of C1q-binding protein is considered to be an important factor in protection.[33]
After complement activation, insertion of MACs into erythrocytes causes lysis that can release
oxyhemoglobin and other substances postulated to damage smooth-muscle cells. Erythrocyte lysate or some component from it can induce spasm in cerebral vessels.[16,18] Notably, membrane conductance changes in smooth-muscle cells have been reported after exposure to lysate, and these changes in
conductance have been attributed to effects of oxygen free radicals.[38] Despite important progress in understanding the effects of lysate on smooth muscle,[14,40,41,43] the in situ dose response and specific mechanism(s) for effects of lysate remain to be elucidated; links between lysate and specific cellular manifestations of spasm, such as altered protein kinase C activity,[24,31,50] have not been demonstrated. It can be inferred from our data that the conductance changes that we observed were due to MAC
insertions rather than to products of erythrocyte lysis. First, at concentrations known to protect from free radical damage,[38] superoxide dismutase plus catalase provided no substantive protection, a finding that agrees with observations in an experimental model of SAH in vivo.[15] More important, the changes in conductance with aged erythrocytes plus autologous serum were similar to those produced by 1)
activation of complement by zymosan without erythrocytes; 2) heterologous C8- and C9-depleted sera supplemented with the missing complement components; and 3) autologous aged erythrocyte ghosts free of lysate. Together, these findings suggest that the magnitude of the changes in conductance produced with aged intact erythrocytes did not require lysate and that insertion of MACs into the smooth-muscle cells was necessary and sufficient for the changes in conductance of smooth-muscle cells observed with aged erythrocytes plus autologous serum.
Formation of C5b-C9 channels in the plasma membrane can lead to erythrocyte lysis or death of nucleated cells. Less well recognized are a number of important effects of sublytic concentrations of MACs.[21,33] Functional sequelae of the C5b through C9 complex binding to smooth-muscle cells have not been identified, but in myocardial cells, insertion of MACs into the membrane can transiently
augment, in a dose-dependent manner, both basal cytosolic Ca++ concentration and Ca++ transients,
resulting in a temporary increase in contractility.[3] Also, at sublytic concentrations, MAC insertions may be associated with receptor-independent activation of guanine nucleotide-binding regulatory protein,[22] increases in diacylglycerol production,[23] and increases in protein kinase C activity,[4] some of which have been observed in vasospastic vessels.[24,31,50] Other effects of sublytic MAC insertions that may be relevant to vasospasm include activation of second messengers,[4,32] production of arachidonic acid,[5,35] production of leukotriene B4,[34] and membrane depolarization.[39]
The bystander hypothesis represents a new mechanism, in addition to hemolysate, that may account for the induction of cerebral vasospasm. As postulated with hemolysate,[38] the bystander mechanism gives rise to changes in membrane permeability that may initiate smooth-muscle cell damage. We posit that an immunological product, the C5b-C9 channel or MAC, may act as a rogue ion channel admitting
extracellular Ca++ to the smooth-muscle cell to cause contraction, activation of protein kinase C, and
other molecular processes that are ultimately manifested as vasospasm. Notably, MAC channels are not
sensitive to typical Ca++ channel blockers or other pharmacological agents that might turn them off, and
the ionic and other cytoplasmic changes that ensue from MAC insertions into smooth-muscle cells may be difficult to reverse by known treatments aimed at achieving vasorelaxation.[8] Although extensive work has shown important effects of hemolysate on cerebrovascular reactivity and tone, the bystander mechanism would still precede release of hemolysate and, thereby, could initiate or contribute to
vasospasm. The two mechanisms are closely intertwined at their origin and determining the relative role of each remains an important challenge. Both account equally well for the timing of vasospasm after hemorrhage, although the bystander hypothesis may, in addition, help explain the development of vasospasm in the absence of hemorrhage.[28,46]
Acknowledgments
We thank Ms. Lioudmila Melnitchenko for technical assistance and Ms. Amy Yerkes for editorial assistance.
References
1. Aoki T, Takenaka K, Suzuki S, et al: The role of hemolysate in the facilitation of
oxyhemoglobin-induced contraction in rabbit basilar arteries. J Neurosurg 81:261-266, 1994 2. Benesch RE, Benesch R, Yung S: Equations for the spectrophotometric analysis of hemoglobin mixtures. Anal Biochem 55:245-248, 1973
3. Berger HJ, Taratuska A, Smith TW, et al: Activated complement directly modifies the performance of isolated heart muscle cells from guinea pig and rat. Am J Physiol 265:H267-H272, 1993
4. Carney DF, Lang TJ, Shin ML: Multiple signal messengers generated by terminal complement complexes and their role in terminal complement complex elimination. J Immunol 145:623-629, 1990 5. Cybulsky AV: Release of arachidonic acid by complement C5b-9 complex in glomerular epithelial
cells. Am J Physiol 261:F427-F436, 1991
6. Fujita T, Takiuchi M, Matsumoto M, et al: C3 and C5-cleaving properdin enzymes formed on
zymosan incubated with human serum: the decay and the regeneration of the enzymes. Int Arch Allergy
Appl Immunol 53:303-309, 1977
7. German JW, Gross CE, Giclas P, et al: Systemic complement depletion inhibits experimental cerebral vasospasm. Neurosurgery 39:141-146, 1996
8. Halperin JA, Taratuska A, Rynkiewicz M, et al: Transient changes in erythrocyte membrane
permeability are induced by sublytic amounts of the complement membrane attack complex (C5b-9).
Blood 81:200-205, 1993
9. Hamill OP, Marty A, Neher E, et al: Improved patch-clamp techniques for high-resolution current recording from cells and cell-free membrane patches. Pflugers Arch 391:85-100, 1981
10. Handa Y, Kabuto M, Kobayashi H, et al: The correlation between immunological reaction in the arterial wall and the time course of the development of cerebral vasospasm in a primate model.
Neurosurgery 28:542-549, 1991
11. Hoshi T, Shimizu T, Kito K, et al: [Immunological study of late cerebral vasospasm in subarachnoid hemorrhage. Detection of immunoglobulins, C3, and fibrinogen in cerebral arterial walls by
immunofluorescence method.] Neurol Med Chir 24:647-654, 1984 (Jpn)
12. Kasuya H, Shimizu T: Activated complement components C3a and C4a in cerebrospinal fluid and plasma following subarachnoid hemorrhage. J Neurosurg 71:741-746, 1989
13. Liem LK, Simard JM, Song Y, et al: The patch clamp technique. Neurosurgery 36:382-392, 1995 14. Macdonald RL, Weir BKA: A review of hemoglobin and the pathogenesis of cerebral vasospasm.
Stroke 22:971-982, 1991
15. Macdonald RL, Weir BKA, Runzer TD, et al: Effect of intrathecal superoxide dismutase and catalase on oxyhemoglobin-induced vasospasm in monkeys. Neurosurgery 30:529-539, 1992
16. Macdonald RL, Weir BKA, Runzer TD, et al: Etiology of cerebral vasospasm in primates. J
Neurosurg 75:415-424, 1991
17. Mayberg MR (ed): Cerebral Vasospasm. Neurosurgery Clinics of North America. Philadelphia: WB Saunders, 1990
18. Mayberg MR, Okada T, Bark DH: The role of hemoglobin in arterial narrowing after subarachnoid hemorrhage. J Neurosurg 72:634-640, 1990
19. Morgan BP: Mechanisms of tissue damage by the membrane attack complex of complement.
Complement Inflamm 6:104-111, 1989
20. Morgan BP: Physiology and pathophysiology of complement: progress and trends. Crit Rev Clin
Lab Sci 32:265-298, 1995
complement membrane attack. Cell Calcium 7:399-411, 1986
22. Niculescu F, Rus H, Shin ML: Receptor-independent activation of guanine nucleotide-binding regulatory proteins by terminal complement complexes. J Biol Chem 269:4417-4423, 1994
23. Niculescu F, Rus H, Shin S, et al: Generation of diacylglycerol and ceramide during homologous complement activation. J Immunol 150:214-224, 1993
24. Nishizawa S, Nezu N, Uemura K: Direct evidence for a key role of protein kinase C in the development of vasospasm after subarachnoid hemorrhage. J Neurosurg 76:635-639, 1992
25. Østergaard JR, Kristensen BØ, Svehag SE, et al: Immune complexes and complement activation following rupture of intracranial saccular aneurysms. J Neurosurg 66:891-897, 1987
26. Pellettieri L, Carlson CA, Lindholm L: Is the vasospasm following subarachnoidal hemorrhage an immunoreactive disease? Experientia 37:1170-1171, 1981
27. Pellettieri L, Nilsson B, Carlsson CA, et al: Serum immunocomplexes in patients with subarachnoid hemorrhage. Neurosurgery 19:767-771, 1986
28. Peterson JW, Kwun BD, Hackett JD, et al: The role of inflammation in experimental cerebral vasospasm. J Neurosurg 72:767-774, 1990
29. Peterson JW, Kwun BD, Teramura A, et al: Immunological reaction against the aging human
subarachnoid erythrocyte. A model for the onset of cerebral vasospasm after subarachnoid hemorrhage. J
Neurosurg 71:718-726, 1989
30. Peterson JW, Roussos L, Kwun BD, et al: Evidence of the role of hemolysis in experimental cerebral vasospasm. J Neurosurg 72:775-781, 1990
31. Sako M, Nishihara J, Ohta S, et al: Role of protein kinase C in the pathogenesis of cerebral vasospasm after subarachnoid hemorrhage. J Cereb Blood Flow Metab 13:247-254, 1993
32. Shin ML, Carney DF: Cytotoxic action and other metabolic consequences of terminal complement proteins. Prog Allergy 40:44-81, 1988
33. Shin ML, Rus HG, Niculescu FI: Membrane attack by complement: assembly and biology of terminal complement complexes. Biomembranes 4:123-149, 1996
34. Shirazi Y, Imagawa DK, Shin ML: Release of leukotriene B4 from sublethally injured
oligodendrocytes by terminal complement complexes. J Neurochem 48:271-278, 1987
35. Shirazi Y, McMorris FA, Shin ML: Mechanism of arachidonic acid release by terminal complement complexes from rat oligodendrocyte X C6 glioma cell hybrids. Ann NY Acad Sci 540:381-383, 1988 36. Shiver JW, Dankert JR, Esser AF: Formation of ion-conducting channels by the membrane attack complex proteins of complement. Biophys J 60:761-769, 1991
37. Simard JM: Calcium channel currents in isolated smooth-muscle cells from the basilar artery of the guinea pig. Pflugers Arch 417:528-536, 1991
cerebrovascular smooth-muscle cells. Circ Res 68:416-423, 1991
39. Stephens CL, Henkart PA: Electrical measurements of complement-mediated membrane damage in cultured nerve and muscle cells. J Immunol 122:455-458, 1979
40. Takanashi Y, Fujitsu K, Fujii S, et al: Altered reactivity of hemolysate-treated cultured
smooth-muscle cells from rabbit basilar artery determined by digital imaging microscopy. J Neurosurg
75:82-90, 1991
41. Takanashi Y, Weir BKA, Vollrath B, et al: Time course of changes in concentration of intracellular free calcium in cultured cerebrovascular smooth-muscle cells exposed to oxyhemoglobin. Neurosurgery
30:346-350, 1992
42. Tewari K, Simard JM: Protein kinase A increases availability of calcium channels in smooth-muscle cells from guinea pig basilar artery. Pflugers Arch 428:9-16, 1994
43. Vollrath BAM, Weir BKA, Macdonald RL, et al: Intracellular mechanisms involved in the responses of cerebrovascular smooth-muscle cells to hemoglobin. J Neurosurg 80:261-268, 1994
44. West GA, Leppla DC, Simard JM: Effects of external pH on ionic currents in smooth-muscle cells from the basilar artery of the guinea pig. Circ Res 71:201-209, 1992
45. Yanamoto H, Kikuchi H, Ishikawa J, et al: Intravenous FUT-175 inhibits complement activation in the cerebrospinal fluid and vasospasm-related delayed ischemic neurological deficit following
subarachnoid hemorrhage, in Findlay JM (ed): Cerebral Vasospasm. New York: Elsevier, 1993, pp 431-434
46. Yanamoto H, Kikuchi H, Okamoto S, et al: Cerebral vasospasm caused by cisternal injection of polystyrene latex beads in rabbits is inhibited by a serine protease inhibitor. Surg Neurol 42:374-381, 1994
47. Yanamoto H, Kikuchi H, Okamoto S: Effects of protease inhibitor and immunosuppressant on cerebral vasospasm after subarachnoid hemorrhage in rabbits. Surg Neurol 42:382-387, 1994
48. Yanamoto H, Kikuchi H, Okamoto S, et al: Preventive effect of synthetic serine protease inhibitor, FUT-175, on cerebral vasospasm in rabbits. Neurosurgery 30:351-357, 1992
49. Yanamoto H, Kikuchi H, Sato M, et al: Therapeutic trial of cerebral vasospasm with the serine protease inhibitor, FUT-175, administered in the acute stage after subarachnoid hemorrhage.
Neurosurgery 30:358-363, 1992
50. Yokota M, Peterson JW, Kaoutzanis MC, et al: Protein kinase C and diacylglycerol content in basilar arteries during experimental cerebral vasospasm in the dog. J Neurosurg 82:834-840, 1995
Manuscript received November 14, 1996. Accepted in final form February 21, 1997.
This manuscript was published previously in J Neurosurg 87:294-300, 1997.
Address reprint requests to: J. Marc Simard, M.D., Ph.D., Department of Neurosurgery, University of Maryland School of Medicine, 22 South Greene Street, Baltimore, Maryland 21201.
