(Stroke. 2000;31:2149.)
© 2000 American Heart Association, Inc.
Original Contributions |
Thrombin Reduces Cerebral Arterial Contractions Caused by Cerebrospinal Fluid From Patients With Subarachnoid Hemorrhage
From the Section of Neurosurgery, Department of Surgery, University of Chicago Medical Center and Pritzker School of Medicine, Chicago, Ill.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
Background and Purpose—We observed that the second application of cerebrospinal fluid (CSF) from subarachnoid hemorrhage (SAH) patients onto cerebral arterial segments in vitro produces a greater contraction than does the initial application. It was hypothesized that the difference between the first and second applications of SAH CSF was due to the activity of thrombin.
Methods—Canine vertebrobasilar artery was removed under general anesthesia, cut into rings, and suspended in tissue culture baths so as to measure isometric tension. CSF was taken from patients 1 to 3 days after SAH via ventricular drains. CSF was administered in 10-5 to 10-1 dilutions. The thrombin antagonist hirudin (5 U) was administered before CSF in some experiments. The arterial tension response to pure oxyhemoglobin (10-4 to 3.2 g/dL) and thrombin (10-4 to 3.2 U/mL), administered alone or in combination, was also examined.
Results—Hirudin increased arterial tension generated on the initial application of SAH CSF but had no effect on the tension generated by the second application of the SAH CSF, suggesting that thrombin limits the tension generated by vasoconstrictive agents in the CSF. Thrombin and pure oxyhemoglobin administered together produced less tension than that generated in response to oxyhemoglobin administered alone; no additive response was observed by coadministering the 2 vasoconstrictive agents.
Conclusions—In the presence of oxyhemoglobin, thrombin acts to reduce cerebral arterial tension. This interaction between thrombin and hemoglobin may account for the observation that the second application of CSF from SAH patients onto cerebral arterial segments in vitro produces a greater contraction than does the initial application.
Key Words: hemoglobin • plasmin • subarachnoid hemorrhage • thrombin • vasospasm
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Recent studies from our laboratory have focused on identifying compounds in the cerebrospinal fluid (CSF) of subarachnoid hemorrhage (SAH) patients that cause contraction of cerebral arteries. These CSF factors are thought to reflect the environment in the subarachnoid clot that contributes to the development of vasospasm. This report describes the importance of thrombin in the generation of cerebral arterial vasoconstriction caused by SAH CSF.
It is generally thought that hemoglobin is the principal spasmogenic agent in the subarachnoid clot.1 Hemoglobin, particularly oxyhemoglobin, is a vasoconstrictive agent both in vitro2 and in vivo.3 Hemoglobin is released during the breakdown of erythrocytes, which begins shortly before the time at which vasospasm develops.4 Although the hemoglobin concentration in erythrocyte hemolysate correlates well with its vasoconstrictive ability,5 the hemoglobin concentration in CSF from SAH patients correlates poorly with the ability of the CSF to cause vasoconstriction.6 The inability of hemoglobin to determine SAH CSF contractility and the disparity in contraction caused by SAH CSF and erythrocyte hemolysate may be due to the presence of vasoactive plasma proteins that would be present in the CSF after SAH.7 8
One such protein is the serine protease thrombin. Thrombin has an established role in vasospasm at sites of blood coagulation in coronary arteries.9 On cerebral arteries, thrombin exerts a biphasic effect, causing transient endothelium-dependent relaxation8 and more prolonged vasoconstriction by a direct action on smooth muscle cells.10 Thrombin in the CSF, as measured by the level of fibrinopeptide A11 or thrombin-antithrombin,12 is predictive of vasospasm, and serine protease inhibitors that antagonize thrombin reduce vasospasm in a dose-dependent manner after intracisternal injection of whole blood.13 14 However, a more specific examination of the function of thrombin in SAH vasospasm with the use of selective antagonists has not been performed.
We became interested in thrombin because of the observation that the first administration of a sample of SAH CSF onto cerebral arterial segments ex vivo produced a smaller increase in tension than did the subsequent administration of the same CSF sample. The data in the present study suggest that the lower tension generated on the initial exposure to SAH CSF is due to thrombin activity, because the initial response could be increased to the level attained by the second SAH CSF exposure if the arterial segment was pretreated with the thrombin antagonist hirudin. Despite its action as a vasoconstrictor, thrombin was also shown to reduce the tension generated by oxyhemoglobin. The ability of thrombin to reduce oxyhemoglobin contractility appears selective, inasmuch as thrombin did not reduce endothelin-1–induced or KCl-induced contractions. No such relation was observed between plasmin and oxyhemoglobin, further indicating the specific nature of the action of thrombin against oxyhemoglobin contraction.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Animal Handling, Cerebral Artery Harvesting, and Measurement of Cerebral Arterial Tension
All procedures on animals were approved by the Institutional Animal Care and Use Committee. The vertebrobasilar arteries of mongrel dogs were prepared as described elsewhere.15 Unless otherwise noted, arterial segments were used in only one experiment. Briefly, 4-mm sections of vertebrobasilar artery were cleaned of branches and connective tissue and were hung under 0.75 g of isometric tension by means of stainless-steel hooks passed through the lumen. Artery segments were hung in Krebs’ buffer–filled organ baths that were heated to 37°C and bubbled with 95% oxygen/5% carbon dioxide. During a 1-hour acclimation period, the organ bath buffer was changed every 20 minutes, at which time the resting tension was reset to 0.75 g. After the 1-hour acclimation period, the resting tension was no longer adjusted, and the artery segments were subject to experimental manipulation.
Oxyhemoglobin, thrombin, and plasmin were added directly into the organ bath in amounts that increased the concentration of the compound in the organ bath in half logarithmic steps. A drop of antifoam B was administered to prevent foaming of the solution. The tension generated in response to a certain dose of a compound was measured 3 minutes after administration of that dose. In experiments in which 2 compounds were coadministered, a solution with both compounds was made beforehand that contained sufficient concentrations of each compound so that the volume of solution administered into the organ bath to obtain a given concentration of each compound was not different from the volume that would have been administered if each compound had been applied separately.
Acquisition and Preparation of Human SAH CSF Samples
All procedures involving humans were carried out under protocols
approved by the Institutional Review Board. Samples of CSF from
patients with SAH were obtained in cases with radiological evidence of
SAH and a known date of hemorrhage onset. In such patients, CSF
was drained from a ventricular catheter into a reservoir
that was emptied by the nursing staff every 8 hours. CSF samples were
collected from the reservoir on either posthemorrhage day 1
(n=2 samples) or on posthemorrhage day 3 (n=3 samples). Each
sample was taken from a different patient, and no patient was treated
with fibrinolytic compounds. CSF samples were individually stored
without any treatment at -80°C and were thawed immediately before
use. CSF samples were diluted in Krebs’ buffer for use in
dose-response experiments. Each sample was individually applied to at
least 2 separate cerebral artery segments.
Acquisition and Preparation of Monkey Subarachnoid
Blood Clot
Measuring the time course of total serine protease activity and
specific thrombin activity in subarachnoid blood clots from
monkeys offered certain advantages over human CSF samples: (1) the
protease activity in the monkey blood clot directly reflects the
microenvironment around the cerebral artery, whereas the protease
activity in the CSF is assumed to be proportionate to that of the blood
clot; (2) in monkeys, the blood clot can be removed after an exact
duration from the subarachnoid space, whereas in SAH patients,
the exact onset of hemorrhage may be known less accurately; and
(3) comorbidities, drug treatments, and the underlying pathology of the
SAH are not confounding issues in the monkey. Subarachnoid
hemorrhage was simulated in monkeys by placing a 10-mL
autologous blood clot unilaterally into the sylvian fissure. One to 7
days later, the animals were euthanized, and the remaining
subarachnoid clot was recovered. Clots were stored at -80°C
until use. The clot was suspended at 1 mg/150 µL of a calcium buffer
at pH 7.4 containing the following (mmol/L): sodium 145, potassium 3.0,
calcium 2.0, magnesium 1.0, chloride 154, glucose 10, and HEPES 10.
After homogenization, the samples were
centrifuged at 3000g for 10 minutes at 4°C. The
supernatant fluids were carefully removed, diluted 1:10 with calcium
buffer, and stored at -80°C until use.
Measurement of Thrombin Enzymatic Activity in Human SAH CSF Samples
and Monkey Subarachnoid Blood Clot
Thrombin activity was measured in the human CSF samples and in
blood clots removed after a period of incubation from the
subarachnoid space of monkeys. The amount of protease activity
was measured by a chromogenic assay with the synthetic
peptide
N-p-tosyl-Gly-Pro-Arg-p-nitroanilide
(Cromozyme) dissolved in distilled water to 1 mg/mL. A standard curve
was prepared by using dilutions of a thrombin stock solution (10 U/mL
in Krebs’ buffer). The standard curve was generated as follows: One
hundred microliters of the thrombin dilution was added to 800 µL
Cromozyme in the presence of 100 µL of oxyhemoglobin (1 g/dL). The
standards were agitated for 90 minutes at room temperature, after which
the reaction was terminated with 100 µL of trichloroacetic acid (5
mol/L). After ultracentrifugation for 15 minutes, 850
µL of the supernatant was removed and mixed with 150 µL of 1N NaOH.
The amount of free nitroanilide was measured at 405 nm in a
spectrophotometer. Total protease activity in samples of human CSF or
monkey clot extract was measured in a manner similar to that used for
the dilutions of stock thrombin solution, except that 100 µL of
individual CSF or extract sample was added to the 800 µL Cromozyme
and 100 µL Ringer’s lactate. All thrombin standards and CSF or clot
extract samples were run in triplicate.
The degree to which the conversion of Cromozyme to free nitroanilide was attributable to thrombin was measured by the addition of excess hirudin (50 U). This amount of hirudin was sufficient to inhibit >99% of the activity of 0.1 U purified thrombin. The 100 µL of lactated Ringer’s solution that was added to the Cromozyme and the CSF or clot extract sample was replaced with 100 µL of hirudin solution; otherwise, the assay was performed as described above. The amount of Cromozyme cleavage in the presence of hirudin minus the amount of Cromozyme cleavage in the absence of hirudin was taken as the amount of specific thrombin activity.
Measurement of Hemoglobin in Human SAH CSF Samples and Monkey
Subarachnoid Blood Clot
The hemoglobin content of CSF samples from patients with SAH or
of subarachnoid clot extracts from monkeys was determined by
the characteristic absorbencies of hemoglobin at 577 and 630 nm
(A577nm and ·
A630nm, respectively). The amounts of hemoglobin
in ferrous and ferric forms (in µmol/L) were then calculated
with the following
equations: [oxyhemoglobin]=[(66 · A577nm)-(80 · A630nm)]/4
 |
). This was necessary because hemoglobin
exhibits a pronounced absorption curve overlying the 405-nm wavelength
(the Soret band) that would interfere with the measurement of free
nitroanilide. Specific measurement of hemoglobin by its absorbance at
577 and 630 nm failed to detect any residual hemoglobin in the
supernatant fluids from the monkey subarachnoid blood clots
after the acid treatment.
|
Drugs and Proteins
Thrombin, plasmin, Cromozyme, antifoam B, inorganic ions, acids,
and bases were purchased from Sigma Biochemicals. Thrombin, plasmin,
and hirudin were purchased in lyophilized form and were dissolved in
Krebs’ buffer immediately before use. Cromozyme was purchased in
lyophilized form and was diluted with distilled water to a stock
concentration of 1 mg/mL.
Hirudin was purchased from American Diagnostica Inc. Hirudin is a well-described inhibitor of thrombin (see References 16 and 19). There are reports that hirudin inhibits the enzymatic activity of plasmin17 and factor IX18 ; however, these have been contradicted (thrombin,16 factor IX20 ). Hirudin has been shown not to inhibit a variety of other proteases.21 22 23 24 25 26
Purified human hemoglobin was the generous gift of Hemosol Inc. Spectrophotometric assessment of the hemoglobin stock solution showed that it was purely in the oxyhemoglobin form. Purified human hemoglobin was supplied at a concentration of 10 g/dL in Ringer’s lactate buffer.
Statistical Analysis
Determination of the response of cerebral artery segments to a
treatment that involves multiple doses or comparison of the response of
the cerebral artery segments to 2 different treatments, each involving
multiple doses, was made by ANOVA. ANOVA results are reported as F and
probability values. Post hoc analysis was performed with the
Tukey methodology. Comparison of the effects of a single treatment (ie,
hirudin treatment) on cerebral artery segment tension was performed
with repeated-measures t tests. The t test
results are reported as t and probability values. A value of
P<0.05 was considered statistically significant.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Hemoglobin Content and Serine Protease Activity of Human SAH CSF Samples
The hemoglobin content of the CSF samples from patients with SAH was measured by spectrophotometry. For the samples (n=5) that were used in cerebral artery tension experiments, the following values were obtained: 27±11 µmol/L oxyhemoglobin and 18±5 µmol/L methemoglobin.
The amount of serine protease activity in the SAH CSF samples was also measured. Specific thrombin activity in the CSF samples was quantified by inhibiting the thrombin in the solution with excess hirudin. The average total protease activity in the CSF samples was found to be 0.014±0.011 U/mL. Only 3 of the CSF samples had sufficient volume remaining for the measurement of specific thrombin activity. In these 3 samples, 34±7% of the total protease activity was inhibited by 50 U hirudin.
Effects of Repeated Administration of Human SAH CSF on Tension
Generated by Cerebral Artery and Role of Thrombin
It was observed that the tension generated in canine
vertebrobasilar arterial segments in response to
application of a sample of SAH CSF was greater if the
arterial segment had been exposed previously to that sample
of CSF. Both the first and second administrations of CSF caused
concentration-dependent contractions (Figure 2; first administration, F=3.8 and
P<0.01; second administration, F=10.3 and
P<0.001). However, tension generated by SAH CSF applied to
a fresh cerebral artery was significantly less than the tension
generated by the same CSF sample applied to a cerebral
arterial segment that had previous been exposed to the CSF
(F=15.7, P<0.001). A significant difference between the 2
groups was noted with as little as a
10-3 dilution of SAH
CSF.
|
This observation led to the hypothesis that the initial exposure to SAH
CSF altered the cerebral arterial responsiveness to some
factor within the CSF in a manner that was not susceptible to washout
and that was long-lived. Hemoglobin was eliminated as this factor
because repeated application of purified oxyhemoglobin does not cause
consecutively greater increases in tension (data not shown). Rather,
the activity of a blood-borne proteolytic enzyme was suspected,
particularly that of thrombin. To test the involvement of thrombin, its
enzymatic activity in the CSF samples was blocked by pretreating the
arterial segments with hirudin, a highly specific
antagonist of thrombin. Figure 3 shows the effect of hirudin on
contractions caused by SAH CSF samples on arteries with or without
prior exposure to SAH CSF. Hirudin by itself had no effect on resting
cerebral artery tension (prehirudin tension 0.64±0.02 g, posthirudin
tension 0.63±0.02 g; t=0.79, P=0.22). However,
hirudin pretreatment significantly reduced tension generated by the
subsequent administration of SAH CSF onto fresh cerebral
arterial segments (F=13.5, P<0.001).
Significantly greater tension was generated in the presence of hirudin
after administration of as little as
3.2x10-1 dilution of CSF.
In contrast, hirudin did not significantly affect the tension generated
by SAH CSF on arterial segments that had been previously
treated with the CSF (F=1.4, P=0.23).
|
Effects of Thrombin on Oxyhemoglobin Vasoconstriction
Responses of arterial segments to purified thrombin
were assessed. Thrombin caused concentration-dependent contractions
(Figure 4, F=4.5,
P<0.001) that could be antagonized by pretreatment with 5 U
hirudin (F=0.2 P=0.97; data not shown). Thrombin in our
experiment did not induce vasorelaxation; this is likely because of the
lack of functional endothelium in our
arterial segments, as was demonstrated by lack of
relaxation in response to ATP
(10-5 mol/L) administered
after contracting the arterial segments with KCl (45
mmol/L) (data not shown).
|
Because oxyhemoglobin is believed to be an important vasoconstrictor in
SAH CSF, the effects of thrombin, oxyhemoglobin, and thrombin plus
oxyhemoglobin on tension generated by arterial segments
were compared (Figure 4
). The effect of thrombin is described
above. Oxyhemoglobin administered alone caused concentration-dependent
increases in tension (F=3.9, P<0.001). Combined
administration of thrombin plus oxyhemoglobin also caused
concentration-dependent increases tension (F=12.5,
P<0.001), but the response was significantly different from
that of either thrombin or oxyhemoglobin administered alone (F=1.8 and
P<0.001 versus thrombin alone, F=13.9 and
P<0.001 versus oxyhemoglobin alone). Tension generated by
arterial segments exposed to oxyhemoglobin plus thrombin
was less than that generated by oxyhemoglobin alone over thrombin
concentrations from 10-4
to 10-2 U/mL, whereas at
the highest doses of thrombin and oxyhemoglobin, the contraction was
greater than that generated by oxyhemoglobin alone.
Measurement of Thrombin Enzymatic Activity in Monkey
Subarachnoid Blood Clot
The enzymatic activity of thrombin in monkey subarachnoid
blood clots was measured over time to ascertain at what times after SAH
thrombin might be acting to limit vasoconstriction. The assay
demonstrated that the total protease activity in the clot decreased
very rapidly during the first 3 days after placement of the blood clot
into the subarachnoid space (Figure 5). Hirudin-sensitive
chromogenic activity (eg, thrombin activity) accounted for
46% of the total protease activity 1 day after placement of the
clot. This result is similar to the fraction of thrombin activity in a
fresh blood clot that was not placed in the subarachnoid space
(48%), represented by the day-0 time point. Thrombin
activity appears to decay more rapidly than does total protease
activity, inasmuch as excess hirudin inhibited only 17% of the total
protease activity in clots removed after 5 to 7 days. By 7 days after
clot placement, this is equivalent to 0.002 U/mL thrombin activity in
the clot.
|
Effect of Thrombin on Tension Generated by Other
Vasoconstrictors
The specificity of the interaction between thrombin and
oxyhemoglobin was examined by assessing the ability of thrombin to
alter the contraction generated by KCl (60 mmol/L) and
endothelin-1 (2x10-11
mol/L). Arterial segments were exposed to thrombin (0.32
U/mL), and after achievement of maximal contraction, either KCl or
endothelin-1 was administered without washing out the thrombin. These
contractions were compared with those caused by KCl or endothelin-1 in
the absence of thrombin precontraction (Figure 6). Pretreatment with thrombin did not
affect the magnitude of contraction to KCl (t=0.16,
P=0.87) or endothelin-1 (t=0.94,
P=0.37). Thus, it is unlikely that the action of thrombin in
reducing the contraction caused by oxyhemoglobin is a general effect of
thrombin on all vasoconstrictive substances.
|
Effects of Plasmin on Oxyhemoglobin Vasoconstriction
To further study the nature of the interaction between
oxyhemoglobin and thrombin, tension responses to oxyhemoglobin in the
presence of plasmin were examined. Plasmin alone did not affect
cerebral arterial tension (Figure 7; F=0.06, P=0.99).
Furthermore, plasmin did not affect the contractile response to
oxyhemoglobin. Rather, oxyhemoglobin coadministered with plasmin caused
a dose-related increase in arterial tension (F=6.7,
P<0.001) that (at the 0.32 and 1 U/mL doses) was greater
than the response in the absence of plasmin (F=8.43,
P<0.001).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
These experiments were initiated by the observation that some factor exists in CSF samples from SAH patients that reduces the contractile activity of CSF and that this relaxant factor is active only during the first exposure of a cerebral artery to CSF. Several observations have suggested thrombin as a potential candidate for the relaxant factor. Thrombin is present in the CSF in proportion to the severity of SAH.11 Thrombin relaxes cerebral arteries through an endothelium-dependent mechanism,8 and its receptor is activated only once by proteolytic cleavage.27 Furthermore, thrombin receptors have been identified in cerebral arteries.28 Therefore, we determined thrombin vasoactivity in SAH CSF by using the specific thrombin antagonist hirudin. Pretreatment with hirudin significantly increased the tension generated by the first administration of SAH CSF, suggesting that thrombin was acting on the naive artery as a vasorelaxant. However, hirudin did not affect the tension generated by the second administration of SAH CSF onto the arterial segment, indicating that there was no vasorelaxing thrombin activity after the first administration of SAH CSF. The simplest explanation for this difference is that thrombin relaxes the arterial segments during the first administration by proteolytic activation of its membrane receptors, which then cannot be replaced with fresh receptors before the second administration of the CSF. An alternative explanation is that the CSF contains a vasoconstrictive substance to which the arterial segments contract more vigorously on successive administrations. The major SAH CSF vasoconstrictive agent is hemoglobin,1 which, we have found, does not cause successively larger contractions of dog or monkey cerebral arterial segments (authors’ unpublished data, 1998).
Because oxyhemoglobin is believed to be a key vasoactive constituent of
CSF after SAH, we tested the effects of thrombin on contractions to
pure oxyhemoglobin. Thrombin was found to significantly reduce the
contractions generated by oxyhemoglobin when the 2 substances were
administered together. The highest doses of thrombin and oxyhemoglobin
did appear to have additive vasoconstrictive effects,
unlike the lower doses. However, the lower doses of thrombin are more
physiologically relevant. This was surprising
because in our experiments, thrombin caused only
concentration-dependent increases in tension. In the monkey
subarachnoid clot samples, levels of thrombin decrease rapidly
after blood clot placement, but there still should be sufficient
thrombin in the clot after 7 days
(2x10-3 U/mL) to reduce
the contraction generated by oxyhemoglobin. Furthermore, the total
hemoglobin content (45 µmol/L or
2.9x10-3 g/mL) and the
specific thrombin activity level
(5x10-3 U/mL) of our
SAH CSF samples would suggest that these compounds relate in vivo as
did the lower concentrations of purified oxyhemoglobin and
thrombin.
The aforementioned relationship between thrombin and oxyhemoglobin appears to be specific. The reduction of oxyhemoglobin vasoconstriction was selective for oxyhemoglobin because thrombin did not reduce the tension generated by other vasoconstrictor agents. Furthermore, the serine protease plasmin did not reduce oxyhemoglobin-induced contraction of cerebral arterial segments. In fact, high concentrations of plasmin potentiated oxyhemoglobin-induced contraction. Plasmin has been reported to enzymatically cleave protease-activated receptor-1, a thrombin receptor,29 and to cause desensitization of biological responses to thrombin.30 Because plasmin did not cause any contraction by itself in our hands (see Reference 8 for a contrary finding), we are left with the possibility that the augmentation of oxyhemoglobin contractility by plasmin is the result of an elimination of an endogenous thrombin vasorelaxing mechanism that is activated by high concentrations of oxyhemoglobin.
The observations in the present study appear to reflect the effects of oxyhemoglobin and thrombin directly on the smooth muscle of the cerebral artery. We used a preparation of cerebral artery that is deficient in endothelium so that thrombin would not be able to reduce artery tension via its well-described endothelium-dependent mechanism.8 Additionally, the absence of a functional endothelium more closely represents a vasospastic artery, which has been shown to have reduced endothelium-dependent vasorelaxant ability.31 Recent studies suggests that the influx of extracellular calcium is an important component in the contraction of smooth muscle caused by oxyhemoglobin.32 On the other hand, thrombin receptors within cerebral artery smooth muscle cells open membrane calcium channels to cause vasoconstriction.10 How such mechanisms might interact to account for the observations reported in the present study is unknown.
Thrombin has been proposed as a causative agent in vasospasm because of its vasoconstrictive action on cerebral arteries and its presence in the subarachnoid clot.33 34 35 36 By extension, a protective role for thrombin antagonists against the development of vasospasm has been proposed.7 Part of the discrepancy between that hypothesis and the findings in the present study may be the vasoactivity of fibrinopeptides, which are the products of the cleavage of thrombin to fibrinogen. Fibrinopeptides are known to cause vasoconstriction in various peripheral arteries.37 Also, high levels of the breakdown products of fibrin have been associated with vasospasm.38 In fact, evidence has been presented that shows that thrombin facilitates the development of vasospasm only through the production of fibrinopeptides. Arutiunov et al39 demonstrated severe vasoconstriction of canine basilar artery after placement of a fibrinogen-thrombin mix around the artery, but no vasoconstriction was observed if either substance was placed around the artery alone. Thus, the association of thrombin with the development of vasospasm may be a spurious one, and instead, we propose that the products of thrombin enzymatic activity promote vasospasm rather than thrombin per se.
In summary, we propose that thrombin reduces the contraction of cerebral arteries caused by hemoglobin and that this occurs within SAH CSF. This interaction between thrombin and oxyhemoglobin appears to be independent of the endothelium and is specific for the 2 substances. On the basis of the time course of thrombin activity in the subarachnoid clot, we hypothesize that the ability of thrombin to reduce hemoglobin contractility accounts in part for the delayed onset of vasospasm until usually 3 or 4 days after SAH. Furthermore, the potency of this action of thrombin allows for the possibility that thrombin limits the contraction generated by SAH CSF even after vasospasm has developed.
|
Acknowledgments |
|---|
These studies were supported by grants to Drs Macdonald and Weir from the Brain Research Foundation and the National Institutes of Health. We thank Hemosol Inc (Toronto, Canada) for supplying hemoglobin.
|
Footnotes |
|---|
Reprint requests to R. Loch Macdonald, MD, Section of Neurosurgery, MC3026, The University of Chicago Medical Center, 5841 S Maryland Ave, Chicago, IL 60637.
Received April 25, 2000; revision received June 16, 2000; accepted June 16, 2000.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
1. Aoki T, Takenaka K, Suzuki S, Kassell NF, Sagher O, Lee KS. The role of hemolysate in the facilitation of oxyhemoglobin-induced contraction in rabbit basilar arteries. J Neurosurg. 1994;81:261–266.[Medline] [Order article via Infotrieve]
2. Tanishima T, Sano K. Studies on the contractile activity of hemoglobin in isolated dog basilar arteries: its relevance to cerebral vasospasms. Neurol Med Chir (Tokyo). 1980;20:481–487.[Medline] [Order article via Infotrieve]
3. Byrne JV, Griffith TM, Edwards DH, Harrison TJ, Johnston KR. Investigation of the vasoconstrictor action of subarachnoid haemoglobin in the pig cerebral circulation in vivo. Br J Pharmacol. 1989;97:669–674.[Medline] [Order article via Infotrieve]
4. Pluta RM, Afshar JK, Boock RJ, Oldfield EH. Temporal changes in perivascular concentrations of oxyhemoglobin, deoxyhemoglobin, and methemoglobin after subarachnoid hemorrhage. J Neurosurg. 1998;88:557–561.[Medline] [Order article via Infotrieve]
5.
Okwuasaba F, Cook D, Weir B. Changes in vasoactive
properties of blood products with time and attempted identification
of the spasmogens. Stroke. 1981;12:775–780.
6. White RP, Macleod RM, Muhlbauer MS. Evaluation of the role hemoglobin in cerebrospinal fluid plays in producing contractions of cerebral arteries. Surg Neurol. 1987;27:237–342.[Medline] [Order article via Infotrieve]
7.
White RP. Vasodilator proteins: role in delayed
cerebral vasospasm. Stroke. 1986;17:207–213.
8. White RP. Comparison of the inhibitory effects of antithrombin III, alpha 2-macroglobulin, and thrombin in human basilar arteries: relevance to cerebral vasospasm. J Cereb Blood Flow Metab. 1987;7:68–73.[Medline] [Order article via Infotrieve]
9. Haver VM, Namm DH. Generation of a vasoactive substance in human plasma during coagulation: evidence of thrombin-induced contraction of rabbit aorta and dog coronary artery. Blood Vessels. 1983;20:92–98.[Medline] [Order article via Infotrieve]
10. White RP, Cunningham MP, Robertson JT. Effect of the calcium antagonist nimodipine on contractile responses of isolated canine basilar arteries induced by serotonin, prostaglandin F2 alpha, thrombin, and whole blood. Neurosurgery. 1982;10:344–348.[Medline] [Order article via Infotrieve]
11. Kasuya H, Shimizu T, Takakura K. Thrombin activity in CSF after SAH is correlated with the degree of SAH the persistence of subarachnoid clot and the development of vasospasm. Acta Neurochir (Wien). 1998;140:579–584.[Medline] [Order article via Infotrieve]
12. Ikeda K, Asakura H, Futami K, Yamashita J. Coagulative and fibrinolytic activation in cerebrospinal fluid and plasma after subarachnoid hemorrhage. Neurosurgery. 1997;41:344–349 [discussion 349–350].[Medline] [Order article via Infotrieve]
13. Aoyama T, Ino Y, Ozeki M, Oda M, Sato T, Koshiyama Y, Suzuki S, Fujita M. Pharmacological studies of FUT-175, nafamstat mesilate, I: inhibition of protease activity in in vitro and in vivo experiments. Jpn J Pharmacol. 1984;35:203–227.[Medline] [Order article via Infotrieve]
14. Yanamoto H, Kikuchi H, Okamoto S, Nozaki K. Preventive effect of synthetic serine protease inhibitor, FUT-175, on cerebral vasospasm in rabbits. Neurosurgery.. 1992;30:351–356 [discussion 356–357].[Medline] [Order article via Infotrieve]
15. Zhang J, Sima B, Johns L, Macdonald R. Time course of changes in arterial relaxation following subarachnoid hemorrhage in dogs. Neurol Res. 1996;18:227–232.[Medline] [Order article via Infotrieve]
16. Stürzebecher J, Walsmann P. Structure-activity relationships of recombinant hirudins. Semin Thromb Hemost. 1991;17:94–98.[Medline] [Order article via Infotrieve]
17. Gilboa N, Villannueva GB, Fenton JW II. Inhibition of fibrinolytic enzymes by thrombin inhibitors. Enzyme. 1988;40:144–148.[Medline] [Order article via Infotrieve]
18.
Rick ME. Activation of factor VIII by factor Ixa.
Blood. 1982;60:744–751.
19. De Filippis V, Russo I, Vindigni A, Di Cera E, Salmaso S, Fontana A. Incorporation of noncoded amino acids into the N-terminal domain 1–47 of hirudin yields a highly potent and selective thrombin inhibitor. Protein Sci. 1999;8:2213–2217.[Medline] [Order article via Infotrieve]
20.
Chang J, Jin J, Lollar P, Bode W, Brandstetter H,
Hamaguchi N, Straight DL, Stafford DW. Changing residue 338 in human
factor IX from arginine to alanine causes an increase in catalytic
activity. J Biol Chem. 1998;273:12089–12094.
21.
Powers CA, Nasjletti A. A novel kinin-generating
protease (kininogenase) in the porcine anterior pituitary. J
Biol Chem. 1982;257:5594–5600.
22. Fischman AJ, Wildey GM, Matsueda GR, Margolies MN, Zisfein JB, Homcy CJ, Graham RM. Specificity of serine proteases for cleavage sites on proatrial natriuretic factor. Peptides. 1988;9:1275–1283.[Medline] [Order article via Infotrieve]
23. Simmons WH, Walter R. Carboxamidopeptidase: purification and characterization of a neurohypophyseal hormone inactivating peptidase from toad skin. Biochemistry. 1980;19:39–48.[Medline] [Order article via Infotrieve]
24. Zisfein JB, Graham RM, Dreskin SV, Wildey GM, Fischman AJ, Homcy CJ. Characterization and purification of a protease in serum that cleaves proatrial natriuretic factor (ProANF) to its circulating forms. Biochemistry. 1987;26:8690–8697.[Medline] [Order article via Infotrieve]
25. Falanga A, Consonni R, Ruggenenti P, Barbui T. A cathepsin-like cysteine proteinase proaggregating activity in thrombotic thrombocytopenic purpura. Br J Haematol. 1991;79:474–480.[Medline] [Order article via Infotrieve]
26. ten Cate H, Biemond BJ, Levi M, Wuillemin WA, Bauer KA, Barzegar S, Buller HR, Hack CE, ten Cate JW, Rosenberg RD. Factor XIa induced activation of the intrinsic cascade in vivo. Thromb Haemost. 1996;75:445–449.[Medline] [Order article via Infotrieve]
27. White RP, Robertson JT. Role of plasmin, thrombin, and antithrombin III as etiological factors in delayed cerebral vasospasm. Neurosurgery. 1985;16:27–35.[Medline] [Order article via Infotrieve]
28.
Papadaki M, Ruef J, Nguyen KT, Li F, Patterson C, Eskin
SG, McIntire L, Runge M. Differential regulation of protease
activated receptor-1 and tissue plasminogen
activator expression by shear stress in vascular smooth
muscle cells. Circ Res. 1998;83:1027–1034.
29. Kuliopulos A, Covic L, Seeley SK, Sheridan PJ, Helin J, Costello CE. Plasmin desensitization of the PAR1 thrombin receptor: kinetics, sites of truncation, and implications for thrombolytic therapy. Biochemistry. 1999;38:4572–4585.[Medline] [Order article via Infotrieve]
30. Kinlough-Rathbone RL, Perry DW, Rand ML, Packham MA. Pretreatment of human platelets with plasmin inhibits responses to thrombin, but potentiates responses to low concentrations of aggregating agents, including the thrombin receptor activating peptide, SFLLRN. Thromb Haemost. 1997;77:741–747.[Medline] [Order article via Infotrieve]
31. Kim P, Sundt TM Jr, Vanhoutte PM. Alterations in endothelium-dependent responsiveness of the canine basilar artery subarachnoid hemorrhage. J Neurosurg. 1988;69:239–246.[Medline] [Order article via Infotrieve]
32. Vollrath BA, Weir BK, Macdonald RL, Cook DA. Intracellular mechanisms involved in the responses of cerebrovascular smooth-muscle cells to hemoglobin. J Neurosurg. 1994;80:261–268.[Medline] [Order article via Infotrieve]
33. Jerius H, Beall A, Woodrum D, Epstein A, Brophy C. Thrombin-induced vasospasm: cellular signaling mechanisms. Surgery. 1998;123:46–50.[Medline] [Order article via Infotrieve]
34. White RP, Shirasawa Y, Robertson JT. Comparison of responses elicited by alpha-thrombin in isolated canine basilar, coronary, mesenteric, and renal arteries. Blood Vessels. 1984;21:12–22.[Medline] [Order article via Infotrieve]
35.
White RP, Chapleau CE, Dugdale M, Robertson JT.
Cerebral arterial contractions induced by human and bovine
thrombin. Stroke. 1980;11:363–368.
36. Nakamura K, Hatano Y, Mori K. Thrombin-induced vasoconstriction in isolated cerebral arteries and the influence of a synthetic thrombin inhibitor. Thromb Res. 1985;40:715–720.[Medline] [Order article via Infotrieve]
37. Moore PK, Hussaini I, Bhardwaj R. Effect of fibrinopeptides A and B on human and rat platelet aggregation in vitro. Eur J Pharmacol. 1988;153:231–238.[Medline] [Order article via Infotrieve]
38. Guggiari M, Dagreou F, Rivierez M, Mottet P, Gallais S, Philippon J, Viars P. Prediction of cerebral vasospasm value of fibrinogen degradation products (FDP) in the cerebro-spinal fluid (CSF) for prediction of vasospasm following subarachnoid haemorrhage due to a ruptured aneurysm. Acta Neurochir (Wien). 1984;73:25–33.[Medline] [Order article via Infotrieve]
39. Arutiunov A, Baron MA, Dobrovol’skii GF, Kornienko VN, Maiorova NA. Prolonged spasm of the arteries of the base of the brain during experimental fibrinolysis in the subarachnoid space. Vopr Neirokhir. 1975;5:5–11.
This article has been cited by other articles:
 |
|
 |
|
  Z.-D. Zhang, B. Yamini, T. Komuro, S. Ono, L. Johns, L. S. Marton, B. Weir, and R. L. Macdonald Vasospasm in Monkeys Resolves Because of Loss of and Encasement of Subarachnoid Blood Clot Stroke, August 1, 2001; 32(8): 1868 - 1874. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||