(Stroke. 1997;28:2031-2038.)
© 1997 American Heart Association, Inc.
Articles |
From the Departments of Neurological Surgery (E.M.O., M.C.W., R.J.T.) and Molecular Biology and Genetics (D.M.P.) of the Johns Hopkins University School of Medicine, Baltimore, Md, and ICOS Corporation, Bothell, Wash (P.A.H., G.N.D.).
Correspondence to Rafael J. Tamargo, MD, Department of Neurological Surgery, The Johns Hopkins Hospital, Meyer 7-113, 600 N Wolfe St, Baltimore, MD 21287-7713.
| Abstract |
|---|
|
|
|---|
Methods The femoral arteries (n=72) in Sprague-Dawley rats (n=36) were enclosed in latex pouches bilaterally. Autologous blood was injected into the pouch on one side, and saline was injected on the contralateral side. Chronic vessel narrowing was evaluated with the use of 29 rats, which were randomized into one of three groups for intraperitoneal injections: (1) antiICAM-1 MAb (2 mg/kg per dose, n=10), (2) isotype-matched MAb (2 mg/kg per dose, n=9), or (3) saline (n=10), given at 3 hours and 3, 6, and 9 days after blood exposure. These rats were killed 12 days after blood exposure, and femoral artery lumen cross-sectional areas were determined by computerized image analysis. Saturation of ICAM-1 binding sites with this dosing schedule was evaluated by fluorescence-activated cell sorter (FACS) analysis of splenocytes. Immunohistochemical studies with objective cell counts were performed to evaluate macrophage/granulocyte infiltration at 24 hours in 7 rats, comparing antiICAM-1 MAb treatment (n=4) with isotype-matched control MAb (n=3).
Results Animals treated with antiICAM-1 MAb showed a significant inhibition of arterial narrowing at 12 days (P=.0081), with lumen patency of 96.5±5.3% (mean±SEM), compared with 77.3±5.6% for isotype-matched MAb and 72.2±5.3% for saline-treated controls. FACS analysis of splenocytes from animals treated with antiICAM-1 MAb confirmed saturation of ICAM-1 binding sites. Vessels treated with antiICAM-1 MAb showed a significant decrease in inflammatory cell infiltrates, with objective macrophage/granulocyte counts of 31.3±26.6 (mean±SEM) per high-powered field, compared with 171.4±30.7 for isotype-matched control MAb (P=.0027).
Conclusions AntiICAM-1 MAb administered systemically starting 3 hours after blood exposure results in significant inhibition of chronic vasospasm in the rat femoral artery model and is correlated with a reduction in the number of infiltrating macrophages and granulocytes in the periadventitial region of blood-exposed arteries. We conclude that inflammatory changes associated with ICAM-1mediated macrophage and granulocyte migration play an important role in the development of posthemorrhagic chronic vasospasm in this model.
Key Words: aneurysm inflammation subarachnoid hemorrhage vasospasm rats
| Introduction |
|---|
|
|
|---|
The recent characterization of three families of CAMs (immunoglobulin superfamily, integrins, and selectins) and the elucidation of their interactions during inflammation,14 allow for a renewed exploration of the inflammatory features of chronic vasospasm. CAMs are selectively expressed on cell membranes of endothelial cells and leukocytes and regulate the critical cellular interactions of the inflammatory response.15 ICAM-1, a member of the immunoglobulin superfamily of CAMs, is a transmembrane glycoprotein expressed on the surface of endothelial cells early in the response to local tissue injury.16 17 ICAM-1 mediates adhesion of macrophages, granulocytes, and eventually lymphocytes to endothelial cells by binding leukocyte-specific integrins (another subfamily of CAMs), namely, lymphocyte function-associated antigen-1 (LFA-1) and complement receptor 3 (CR3), also called Mac-1.18
We have previously shown that the expression of endothelial ICAM-1 increases 3 to 24 hours after the deposition of blood on the adventitial surface of arteries in the rat femoral artery model of vasospasm and that this increased expression correlates with the development of chronic vasospasm, which peaks 12 days after the hemorrhage.19 Animal studies by other groups have shown increased expression of ICAM-1 in the cerebral vasculature in response to ischemia20 and SAH.21 In humans, increased ICAM-1 expression in the brain has been documented in infarcts, viral encephalitis lesions, and multiple sclerosis plaques.22 Therefore, ICAM-1 appears to play a role in a wide range of CNS pathologies and may be similarly important in chronic vasospasm. In this study we demonstrate the specific inhibition of chronic vasospasm and ICAM-1mediated endothelial cell-leukocyte interaction with the systemic administration of murine monoclonal antibodies against ICAM-1 in the rat femoral artery model of vasospasm.
| Materials and Methods |
|---|
|
|
|---|
|
MAb Treatment
An anti-rat ICAM-1 MAb (27E4B, ICOS Corp) was generated by
immunizing BALB/c mice three times at 21-day intervals with an
activated rat T-cell line. Splenic lymphocytes from a single
mouse were fused with NS-1 myeloma cells, and the resulting hybridomas
were established and screened according to standard procedures.
Supernatants were chosen for cloning based on their ability to block
homotypic aggregation of concanavalin Astimulated spleen cell blasts
in response to PMA. An isotype-matched MAb (1B7.11, ATCC TIB 191,
anti2,4,6-trinitrophenyl) was used as a control.
Twenty-nine rats were used to evaluate the efficacy of antiICAM-1 MAb treatment for the inhibition of chronic vasospasm. These animals were randomized into three groups and received intraperitoneal injections of either antiICAM-1 MAb 27E4B (isotype IgG1, 2 mg/kg per dose, n=10), isotype-matched control MAb 1B7.11 (isotype IgG1, 2 mg/kg per dose, n=9), or saline solution (n=10) at 3 hours, 3 days, 6 days, and 9 days after surgery. The animals were killed 12 days after surgery, and the femoral arteries were harvested for routine histology and for measurement of lumen cross-sectional areas. This time point was chosen based on previous studies indicating that maximal vessel lumen narrowing occurs 12 days after blood exposure in this model.19
Seven rats were used for immunohistochemical studies to evaluate the infiltration of macrophages and granulocytes into the region of the blood-exposed femoral arteries 24 hours after surgery. The 24-hour time point has been previously shown to be within the window of increased ICAM-1 expression in this model.19 These animals received intraperitoneal injections of the antiICAM-1 MAb 27E4B (isotype IgG1, 2 mg/kg per dose, n=4) or the isotype-matched control MAb 1B7.11 (isotype IgG1, 2 mg/kg per dose, n=3) 3 hours after blood injection and were killed 24 hours after surgery. The femoral arteries were harvested for immunohistochemical staining for macrophages and granulocytes.
Blood Vessel Preparation
Femoral arteries were perfusion-fixed in situ with 4%
paraformaldehyde in 0.1 mol/L PBS (pH 7.4) via
cannulation of the aorta. The arteries were dissected out of the latex
pouches and immersion-fixed in 4% paraformaldehyde for
an additional 24 hours at 4°C. After fixation, the vessels were
placed in 20% sucrose at 4°C for 24 hours for cryoprotection and
frozen in tissue-embedding matrix by immersion in dry iceequilibrated
isopentane at -50°C. The embedded vessels were stored at -40°C
until sectioning.
FACS Analysis for Determination of ICAM-1 Binding Site
Saturation
Four rats from the efficacy experiment, 2 treated with the
antiICAM-1 antibody and 2 with the isotype-matched control antibody,
were used to evaluate saturation of ICAM-1 binding sites in treatment
animals. Their spleens were harvested at the 12-day time point and
minced in a sterile Petri dish with RPMI medium. The resulting
splenocyte suspension was washed with RPMI and pelleted by
centrifugation at 1500 rpm at 4°C. The pellet was
then resuspended in ACK lysis buffer (0.15 mol/L
NH4Cl, 1.0 mmol/L KHCO3, 0.1
mmol/L Na2EDTA, pH 7.4) and incubated at room
temperature for 5 minutes to lyse red blood cells. The cells were
washed three times with RPMI and resuspended in flow cytometry wash
buffer (2% calf serum, 10 mmol/L HEPES, and 0.1%
NaN3 in Hanks' balanced salt solution).
For each animal, two aliquots of 1x106 cells were pelleted by centrifugation at 1000 rpm for 5 minutes at 4°C. One of these aliquots from each animal was incubated with additional mouse antiICAM-1 antibody (1 µL, 3.2 mg/mL) for 20 minutes on ice to fully saturate ICAM-1 binding sites before FACS. Both aliquots of cells were then incubated with phycoerythrin-conjugated anti-mouse IgG for 20 minutes on ice. After incubation, the cells were washed and resuspended in 0.5 mL of wash buffer and run on FACScan (Becton Dickinson). ICAM-1 binding site saturation was determined by comparison of FACS histograms for cell samples with and without additional antiICAM-1 antibody.
Measurement of Blood Vessel Cross-Sectional Areas
The vessels were sectioned in a transverse orientation with the
use of a cryostat (Microm) at a thickness of 10 µm and stained
with hematoxylin and eosin. Lumen cross-sectional areas were determined
by computerized image analysis on an Apple Macintosh 8100/80AV
computer with the use of the public domain NIH Image program (developed
at the US National Institutes of Health and available from the Internet
by anonymous FTP from zippy.nimh.nih.gov). To correct for vessel
deformation and off-transverse sectioning, the areas were determined by
measuring the circumference of the vessel lumen and calculating the
area as a generalized circle based on the measured circumference. For
each vessel, five separate sections at least 200 µm apart were
measured and averaged. For comparisons among treatment groups, the
ratios of lumen cross-sectional areas of blood-exposed arteries to
saline-exposed arteries in the same animal were calculated. Comparison
of the group mean ratios was performed by the one-way ANOVA method
(JMP-IN, SAS Institute).
Immunohistochemical Staining for Macrophages and
Granulocytes
Immunohistochemical staining was used to evaluate the
infiltration of macrophages and granulocytes into the
blood-exposed and control femoral arteries. The femoral arteries were
sectioned in a transverse orientation with the use of a cryostat
(Microm) at a thickness of 10 µm, with at least 100 µm
between sections. The slides were hydrated in PBS, then incubated in
3% normal horse serum for 20 minutes. The mouse OX-41 MAb (MAS 369,
Harlan Sera-lab, Ltd) was diluted 1:200 in PBS with 1.0% bovine serum
albumin and 3% normal horse serum, applied to each section,
and incubated for 45 minutes in a humidified chamber. OX-41 detects a
membrane antigen found on rat granulocytes and the majority of
macrophages. It does not label lymphocytes or their precursors,
mast cells, or platelets.24 25 After the slides were
washed with PBS, they were incubated with biotinylated, horse
anti-mouse IgG (rat absorbed, BA-2001, Vector Laboratories) diluted
1:100 in PBS with 1.0% bovine serum albumin and 3% normal
horse serum for 45 minutes. The slides were washed in PBS, then
incubated with avidin and biotinylated horseradish peroxidase
macromolecular complex (Vectastain Elite, Vector Laboratories) for 30
minutes. The slides were washed in PBS and incubated for 8 minutes with
peroxidase substrate (Vector VIP, Vector Laboratories). The slides were
then washed in tap water, counterstained with nuclear fast red, and dry
mounted.
A macrophage and granulocyte count was obtained by identifying all positive staining cells per representative high-powered field in a region adjacent to the adventitia of blood-exposed arteries. For each animal, three high-powered field sections of the blood-exposed artery were counted. Cell counts for antiICAM-1treated and isotype-matched control arteries were compared with a two-group t test (JMP-IN, SAS Institute).
| Results |
|---|
|
|
|---|
FACS analysis performed on splenocytes from animals treated
with 4 intraperitoneal doses of the antiICAM-1
MAb or control antibody (3 hours, 3 days, 6 days, and 9 days after
blood exposure) revealed that ICAM-1 binding sites were saturated.
Splenocytes harvested at day 12 from animals treated with the
isotype-matched control antibody demonstrated a shift in the FACS
histogram after addition of the antiICAM-1 MAb, indicating that
ICAM-1 binding sites were not saturated in the control animals (Fig 2A
). There was no shift in the FACS
histogram from animals treated with the antiICAM-1 antibody when
additional antiICAM-1 MAb was added to the cell preparation before
FACScan (Fig 2B
), confirming that ICAM-1 binding sites were
saturated.
|
Vessel Lumen Narrowing From Blood Exposure
The mean ratios of blood-exposed to saline-exposed lumen
cross-sectional area, expressed as percent lumen patency, were
96.5±5.3% (mean±SEM, n=10) for the animals treated with the
antiICAM-1 MAb compared with 77.3±5.6% (n=9) for the animals
treated with the isotype-matched control MAb and 72.2±5.3% (n=10) for
saline control animals (Fig 3
). This
difference in lumen patency for the antiICAM-1 MAb treatment
represents an 87.4% inhibition of vasospasm compared with
control vessels and is statistically significant by one-way ANOVA
(P=.0081).
|
There was no significant systemic toxicity attributed to the 12-day antiICAM-1 MAb treatment. There was no apparent inhibition of surgical wound healing or increase in surgical wound infections in the animals that received the antiICAM-1 antibody. Histological examination of the blood vessels with hematoxylin and eosin staining did not reveal any deleterious effects of antiICAM-1 treatment compared with vessels of isotype-matched control antibodytreated and saline-treated control animals.
Immunohistochemical Staining for Macrophages and
Granulocytes
Immunohistochemical staining for macrophages and
granulocytes revealed a significant decrease in the inflammatory cell
infiltrate in the region of the periadventitial blood clot in the
animals treated with the antiICAM-1 MAb compared with that of control
animals 24 hours after blood exposure (Fig 4
). In animals treated with antiICAM-1
MAb, macrophage/granulocyte counts were 31.3±26.6 (mean±SEM,
n=12 vessel sections) per high-powered field compared with 171.4±30.7
(n=9 vessel sections) per high-powered field for control animals (Fig 5
). This decrease in the number of
inflammatory cells for the animals treated with antiICAM-1 treated
animals was statistically significant (P=.0027, two-group
t test).
|
|
| Discussion |
|---|
|
|
|---|
In the present study the intraperitoneal administration of the antiICAM-1 MAb 27E4B, starting 3 hours after blood deposition and every 3 days subsequently, resulted in significant inhibition of chronic vasospasm at 12 days. FACS analysis of splenocytes from treated animals confirmed saturation of ICAM-1 binding sites with the dosing regimen used. Immunohistochemical studies demonstrated a significant decrease in the number of macrophages and granulocytes at 24 hours (toward the end of the window of increased endothelial ICAM-1 expression) in the periadventitial region of blood-exposed arteries treated with the antiICAM-1 MAb. These results suggest that after a perivascular hemorrhage, early ICAM-1mediated interactions between endothelial cells and macrophages or granulocytes influence the development of chronic vasospasm 12 days later. Therefore, in this experimental model inflammation appears to play a major causative role in the development of chronic vasospasm.
Although the rat femoral artery is not a CNS vessel, it appears to respond to the adventitial blood-induced injury in a manner similar to that observed in human and primate intracranial vessels, particularly in the time course of delayed luminal narrowing and the histological appearance of the affected vessels. In humans and animals, prolonged exposure to blood after SAH results in two phases of lumen narrowing of the cerebral vessels, described as acute and chronic vasospasm.26 27 28 29 30 Acute or active vasospasm occurs hours after SAH, involves active vasoconstriction, and responds to papaverine. Chronic or passive vasospasm, by contrast, occurs roughly 1 week after SAH, involves passive vasoconstriction, and is insensitive to papaverine. We have identified these two phases in the femoral artery model of vasospasm, with significant peak lumen narrowing occurring between 1 and 12 hours and then again 12 days after the hemorrhage.19 The similarity in the time course of vessel narrowing suggests that the mechanism and pathophysiology of the blood injury in the rat femoral artery may be similar to that in human CNS vessels after SAH and that the findings in the femoral artery model of vasospasm can be cautiously generalized to the intracranial vasculature. Moreover, the histological changes associated with this vessel narrowing in the rat femoral artery model also closely parallel those seen in intracranial vessels in human SAH and animal models of SAH, most notably, thickening and corrugation of the internal elastic lamina, medial thickening, and infiltration of granulocytes and mononuclear cells.1 2 5 23
While cerebral arteries differ from systemic arteries in their endothelial permeability, response to vasoactive pharmacological agents, and components of the adventitial matrix, the similarities described above suggest an inflammatory mechanism that is preserved across both vessel types. Furthermore, the ICAM-1mediated inflammatory mechanism that we consider here has been observed in intracranial arteries in response to both ischemic injury20 and SAH21 in animal models. These findings help validate the femoral artery model, as originally described by Okada and colleagues,23 as an appropriate system in which to study basic molecular mechanisms of inflammation in response to vessel or tissue injury that may be preserved across a wide variety of blood vessel types, such as the endothelial expression of ICAM-1. The morphological and physiological differences between peripheral and CNS arteries, or the concern that the degradation of blood peripherally differs significantly from its degradation in the subarachnoid space, should not invalidate the basic mechanisms of inflammation observed in this model.
Inflammation has long been suspected to play a role in chronic vasospasm. Aneurysmal SAH patients with severe chronic vasospasm have been shown to have significantly higher systemic temperatures than those of SAH patients without vasospasm.31 32 In vessels exposed to blood after SAH, Crompton1 identified collections of granulocytes and mononuclear leukocytes below the endothelial layer, and Hughes and Schianchi2 described numerous macrophages in the media and adventitia. These histological findings have also been observed in the primate SAH model.5 Several investigators have reported increased levels of immunoglobulins and complement fractions in the serum and vessel walls of patients in vasospasm,7 8 9 10 and one group has correlated these increased levels with poor outcome.9 Ibuprofen, a drug that not only disrupts prostaglandin synthesis but also specifically inhibits leukocyte-endothelial cell interactions,33 has been shown to prevent chronic vasospasm in dogs.11 12 These observations provide a body of evidence, though indirect, in favor of an inflammatory etiology in chronic vasospasm.
In the present study we demonstrate a primary role for ICAM-1 in the etiology of posthemorrhagic chronic vasospasm and directly implicate macrophages and granulocytes in this phenomenon. Handa and colleagues21 have shown an increased ICAM-1 expression in intracranial vessels in a rat model of SAH. Although the role of ICAM-1 in chronic vasospasm is only now becoming apparent, its role in cerebral ischemia is well documented. ICAM-1mediated extravasation of leukocytes into ischemic cerebral tissue appears to potentiate the ischemic injury and may play an important role in the reperfusion injury of this tissue.20 34 35 Inhibition of leukocyte-endothelial cell binding using monoclonal antibodies against ICAM-1 results in a significant reduction of ischemic brain injury after temporary vascular occlusion.35 36 37 The potential applications of antiICAM-1 therapy in neurosurgical practice include cerebral protection during temporary clipping in aneurysm surgery, management of stroke, and the prevention of chronic vasospasm.
The role of ICAM-1, moreover, represents only one step for possible study in the pathway of endothelial cellmediated migration of granulocytes (primarily neutrophils) and macrophages (derived from monocytes). The extravasation of leukocytes across the endothelial layer occurs in three distinct stages, with each stage involving a distinct set of CAMs.38 39 The first stage involves expression of selectins on the surface of activated endothelial cells within minutes to hours of injury and results in reversible binding of neutrophils and monocytes to the endothelium. We have focused on the second stage, which is mediated by ICAM-1 expressed on endothelial cells, because it results in irreversible binding of leukocytes that express the integrins LFA-1 and CR3 (Mac-1) to the endothelium. The third stage is mediated by cluster of differentiation 31 (CD31) and results in diapedesis and extravasation of these leukocytes across the endothelium.
The functions of these activated macrophages and neutrophils, however, may not be limited to phagocytosis. Leukocytes can induce inhibition of endothelial-dependent relaxation of the cerebral arteries by producing endothelins, which are powerful vasoconstrictors, and thus may directly affect vascular tone via mechanisms separate from their chronic inflammatory effect.40 In this way the inflammatory pathogenesis of chronic vasospasm can be reconciled with the currently favored hypothesis of nitric oxide/endothelin mediation of vasospasm.41 42 Endothelial cells regulate vascular tone by balancing the production of vasodilatory endothelium-derived relaxing factor, which has recently been identified as nitric oxide, and endothelins, a family of peptides that are potent vasoconstrictors. This endothelial-mediated balance of vascular tone can be disrupted by exogenous endothelin production from activated macrophages43 and neutrophils.44 Furthermore, activated neutrophils can directly inhibit endothelium-dependent relaxation,45 46 and activated macrophages can secrete toxic products implicated in chronic vasospasm, such as hydrogen peroxide, superoxide anion, hydroxyl radical, perhydroxyl radical, and singlet oxygen during enhanced phagocytosis.38 42 47
In summary, we describe the inhibition of posthemorrhagic chronic vasospasm in the rat femoral artery model using a murine antiICAM-1 MAb administered systemically starting 3 hours after the hemorrhage. We demonstrate a reduction in the number of infiltrating macrophages and granulocytes seen in the periadventitial region of blood-exposed arteries of animals treated with antiICAM-1 MAb. These results indicate that ICAM-1mediated macrophage and granulocyte migration into the arterial wall may play a major role in the development of posthemorrhagic chronic vasospasm.
| Selected Abbreviations and Acronyms |
|---|
|
| Acknowledgments |
|---|
Received May 5, 1997; revision received June 16, 1997; accepted June 25, 1997.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
K. L. Chaichana, A. P. Levy, R. Miller-Lotan, S. Shakur, and R. J. Tamargo Haptoglobin 2-2 Genotype Determines Chronic Vasospasm After Experimental Subarachnoid Hemorrhage Stroke, December 1, 2007; 38(12): 3266 - 3271. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. L. Macdonald, Z.-D. Zhang, M. Takahashi, E. Nikitina, J. Young, A. Xie, and L. Larkin Calcium sensitivity of vasospastic basilar artery after experimental subarachnoid hemorrhage Am J Physiol Heart Circ Physiol, June 1, 2006; 290(6): H2329 - H2336. [Abstract] [Full Text] [PDF] |
||||
![]() |
C J M Frijns, R Fijnheer, A Algra, J A van Mourik, J van Gijn, and G J E Rinkel Early circulating levels of endothelial cell activation markers in aneurysmal subarachnoid haemorrhage: associations with cerebral ischaemic events and outcome J. Neurol. Neurosurg. Psychiatry, January 1, 2006; 77(1): 77 - 83. [Abstract] [Full Text] [PDF] |
||||
![]() |
H. Yatsushige, M. Yamaguchi, C. Zhou, J. W. Calvert, and J. H. Zhang Role of c-Jun N-Terminal Kinase in Cerebral Vasospasm After Experimental Subarachnoid Hemorrhage Stroke, July 1, 2005; 36(7): 1538 - 1543. [Abstract] [Full Text] [PDF] |
||||
![]() |
C.J.M. Frijns and L.J. Kappelle Inflammatory Cell Adhesion Molecules in Ischemic Cerebrovascular Disease Stroke, August 1, 2002; 33(8): 2115 - 2122. [Abstract] [Full Text] [PDF] |
||||
![]() |
B. Noel, J. E. Conway, and R. J. Tamargo Vascular Complications of Cocaine Use * Response Stroke, July 1, 2002; 33(7): 1747 - 1748. [Full Text] [PDF] |
||||
![]() |
J J Nissen, D Mantle, B Gregson, and A D Mendelow Serum concentration of adhesion molecules in patients with delayed ischaemic neurological deficit after aneurysmal subarachnoid haemorrhage: the immunoglobulin and selectin superfamilies J. Neurol. Neurosurg. Psychiatry, September 1, 2001; 71(3): 329 - 333. [Abstract] [Full Text] [PDF] |
||||
![]() |
Y. Aihara, H. Kasuya, H. Onda, T. Hori, J. Takeda, and F. M. Faraci Quantitative Analysis of Gene Expressions Related to Inflammation in Canine Spastic Artery After Subarachnoid Hemorrhage Editorial Comment Stroke, January 1, 2001; 32(1): 212 - 217. [Abstract] [Full Text] [PDF] |
||||
![]() |
Q.-A. Thai, E. M. Oshiro, R. J. Tamargo, and J. P. Muizelaar Inhibition of Experimental Vasospasm in Rats With the Periadventitial Administration of Ibuprofen Using Controlled-Release Polymers • Editorial Comment Stroke, January 1, 1999; 30(1): 140 - 147. [Abstract] [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Stroke Home | Subscriptions | Archives | Feedback | Authors | Help | AHA Journals Home | Search Copyright © 1997 American Heart Association, Inc. All rights reserved. Unauthorized use prohibited. |