Neuronal Necrosis After Middle Cerebral Artery Occlusion in Wistar Rats Progresses at Different Time Intervals in the Caudoputamen and the Cortex
Background and Purpose Most brain lesions that develop after an artery is occluded evolve from an initial stage of “ischemic injury” (probably reversible) to an infarct or an area where most neurons become necrotic. There is scant information on the time that must elapse after the arterial occlusion for neurons to undergo irreversible injury. The objective of these experiments was to chart the time course and the topographic distribution of the neuronal necrosis that follows the occlusion of a large cerebral artery.
Methods One hundred fifty-one adult rats (including 15 controls) were used in this study. One hundred forty-seven had the right middle cerebral artery occluded for variable periods ranging from 30 minutes to 7 days. After processing the brains for histology, a meticulous structural evaluation of each specimen, including quantitation of necrotic neurons, was followed by a detailed statistical analysis of the neuronal counts.
Results Few neurons in isolated sites showed morphological signs of necrosis during the initial 4 hours; the first significant increase in the percentage of necrotic neurons (15%) was observed within the territory of the occluded artery after 6 hours (P<.05); 12 hours after the arterial occlusion most neurons (65%) had become necrotic (P<.0001). Pannecrosis involving neurons, glial cells, and blood vessels was observed at 72 to 96 hours. However, even at this time pannecrosis involved only the preoptic area and the lateral putamen; a few intact neurons remained visible in the cortex, and scattered necrotic neurons could be identified beyond the edges of the “area of pallor,” which does not become clearly demarcated until 4 to 5 days after the arterial occlusion.
Conclusions There is a predictable progression in the development of neuronal necrosis after a permanent arterial occlusion. Irreversible changes appear first in the caudoputamen and then spread to the cortex. The causes for the progression of the lesion are not known; however, therapeutic interventions that start within the first 1 to 2 hours after the arterial occlusion may alter the histopathologic responses to this form of injury. It remains to be determined whether the extent of the neurological deficit induced by an arterial occlusion correlates with the number of necrotic neurons.
In baboon experiments of middle cerebral artery (MCA) occlusion a close correlation has been established among local cerebral blood flow (CBF) values, loss of action potentials, and ionic pump failure.1 Isoelectric electroencephalographic tracings correspond to sites where CBF equals 0.18 mL · g−1 · min−1, but these abnormalities are fully reversible by reperfusion.1 Carotid artery occlusion in normothermic humans (incidental to the completion of an endarterectomy) is tolerated by most subjects for periods of up to 30 minutes with eventual disappearance, on reperfusion, of electroencephalographic abnormalities and neurological deficits that develop perioperatively.2 3 Massive escape of intracellular potassium and irreversible injury occur promptly at sites where CBF drops below 0.10 mL · g−1 · min−1.1 Low CBF values characterize the core (or center) of the lesion, where necrosis would be expected to develop first. In contrast, the impairment of the blood flow might be less severe at the margins (penumbra) of the arterial territory. In addition to topographical location, a second factor that influences the development of ischemic necrosis is the time elapsed after the arterial occlusion4 ; this suggests that after a large-artery occlusion there may be a continuing deterioration of local CBF, but the factors responsible for this deterioration are not known.
We hypothesized that after occluding the MCA, neuronal necrosis would advance in a predictable manner, both in terms of topographical areas involved (necrotic neurons would appear in large numbers earlier in the caudoputamen than in the cortex) and as a function of time elapsed after the MCA occlusion. Establishing the tempo and the topographical distribution of neuronal necrosis should provide useful guidelines for the time limits during which therapeutic intervention may be efficacious in this paradigm of brain infarct.
We asked the following questions: (1) In rats subjected to MCA occlusion, when does the number of necrotic neurons peak? (2) Does the process responsible for necrosis affect neurons in various areas of the MCA territory at the same time?
One hundred forty-seven adult rats had the right MCA occluded through the insertion of a monofilament via the external carotid artery5 ; experiments were terminated at variable intervals beginning at 30 minutes and lasting up to 7 days after MCA occlusion. Counts of necrotic neurons at three sites—preoptic area, caudoputamen, and frontoparietal cortex—were carried out up to 24 hours after the arterial occlusion.
The number of necrotic neurons increased most markedly during the interval 6 to 12 hours after the MCA occlusion, but at almost all time intervals necrotic neurons were more abundant in the preoptic area and caudoputamen than in the frontoparietal cortex. This time-dependent difference in the progression of the neuronal ischemic injury suggests that the period available for rescuing cortical neurons from the effects of ischemia may be longer than the interval necessary to salvage the neurons of the caudoputamen.
Materials and Methods
All experiments were conducted according to the guidelines issued by the Institutional Animal Care Committee and were in compliance with regulations formulated by the US Department of Agriculture.6 Outbred Wistar rats were purchased from Charles River Laboratory (Wilmington, Mass) and were quarantined for at least 7 days before the day of the experiment.
This study analyzes the effects that permanent MCA occlusion may have on the progression of neuronal necrosis. Two animals that died spontaneously 3 and 4 days after permanent MCA occlusion were not included in this study. One hundred fifty-one adult rats (body weight, 270 to 290 g) were housed individually and fed Agway rat chow for 7 days. Four rats were used as normal controls, and the remaining 147 were used in experiments based on the surgical insertion of a nylon monofilament (4-0) through the external carotid artery. The length of the filament inserted into the vessel was adjusted according to body weight and varied from 18 to 20 mm; this resulted in occlusion of the ostium of the ipsilateral MCA. Further details of the surgical procedure have been published elsewhere7 8 ; briefly, each rat was fasted overnight and anesthetized with a mixture of halothane and nitrous oxide before inserting a nylon filament through the right external carotid artery. The surgical procedure is an adaptation of the method originally described by Koizumi et al9 and Zea Longa et al5 and in most instances was completed in 20 to 25 minutes.
At the conclusion of the experiment and with the use of analeptics (ketamine and xylazine), brains were fixed by cardiovascular perfusion (at constant pressure of 100 mm Hg) with an aldehyde fixative (4% paraformaldehyde in 0.1 mol/L phosphate buffer). Further details on the brain fixation procedures have been published previously.7 8
After the cardiovascular perfusion each specimen was immersed in fixative overnight; thereafter, each brain was cut into five coronal slabs, each measuring 3.0 mm in thickness; these slabs were labeled A (frontal) through E (occipital). Histology sections, approximately 5.0 μm thick, were stained with hematoxylin-eosin. Slab section C (corresponding to the level of the anterior commissure) was embedded in araldite, cut at 1.0-μm-thick sections, and stained with toluidine blue (n=90) for the purpose of selecting appropriate areas for ultrathin sectioning. Twenty-one samples from randomly selected animals were examined by electron microscopy (Table 1⇓).
The 151 rats used in these experiments were randomly divided into 15 experimental subgroups and two control groups: normal (n=4) and sham operated (n=11) (Table 1⇑). The sham-operated animals were subjected to the same surgical and anesthetic procedures used in the experimental subgroups, but the intravascular filament was withdrawn within less than 1 minute.
Definition of Necrotic Neurons
We applied criteria developed by Farber et al10 11 and Trump et al,12 who outlined morphological features (light and electron microscopy) of cells irreversibly injured by ischemia. By light microscopy necrotic neurons can be identified as exhibiting pyknosis, karyorrhexis, karyolysis, and cytoplasmic eosinophilia or loss of hematoxylin affinity. These features can be encompassed in either of these designations: pyknosis/eosinophilia (red neurons) or complete loss of hematoxylinophilia (ghost neurons) (Fig 1⇓). Other cellular alterations such as dark, scalloped, and swollen neurons (Fig 1⇓) were also included in the neuronal counts. These features are reflected in breaks in plasma or nuclear membranes and electron-dense precipitates (calcium salts) in the inner matrix of the mitochondria.13
Quantitation of Neuronal Alterations
The methods used to quantitate neuronal alterations varied slightly as follows: (1) In the control groups, dark neurons, as identified in hematoxylin-eosin preparations (Fig 1B⇑), were counted in each subject in 20 nonoverlapping microscopic fields at a magnification of ×600. (2) In experimental subgroups that lived 24 hours or less, we counted numbers of intact, dark, scalloped/swollen, and necrotic neurons using both 1.0-μm-thick sections stained with toluidine blue and 6.0-μm-thick sections stained with hematoxylin-eosin. The features defining each of these cell types are illustrated in Fig 1⇑. A Sony computer-controlled display video camera interfaced with an Olympus microscope system (Imagist-2; PGT) was used in these quantitations. Thirty nonoverlapping fields at ×600 magnification and representing the preoptic area, caudoputamen, and frontoparietal cortex were collected from the territory of the occluded MCA. The mean percentage of each neuronal type was calculated at each time interval and at each location. (3) In animals that lived more than 24 hours after MCA occlusion, the number of necrotic neurons was not calculated because in these samples necrotic neurons were too numerous; instead we counted intact neurons. Additionally, in each of the animals included in the experimental groups we calculated the area of pallor (in hematoxylin-eosin preparations) according to methods previously described.7
Nineteen subjects from the experimental subgroups and two subjects from the control groups were randomly chosen to have samples from each of the three areas (preoptic area, caudoputamen, and frontoparietal cortex) examined by electron microscopy. The examination concentrated on evaluating features of neuronal perikarya for the purpose of corroborating the classification made by light microscopy. A total of 385 electron micrographs, or an average of 20.2 per animal, were examined to establish the fine detail of neuronal abnormalities.
Data used in the analysis include those derived from light and electron microscopic evaluations of neuronal abnormalities. Results of individual counts and measurements are expressed as mean group values ±SD. ANOVA followed by Bonferroni corrected t test was applied to determine significant differences between control versus experimental groups and between individual experimental subgroups.
The mortality among the group of 153 animals with permanent MCA occlusion was 1.47%.
Control and Sham-Operated Groups
Most neurons in the brain sections from these groups of animals were of the intact type (Figs 1A⇑ and 2A⇓). Dark neurons (Figs 1B⇑ and 2B⇓) were visible at a rate of almost 11% per microscopic field (Table 2⇓); these dark neurons were present in both hemispheres of the control and the experimental subgroups. A very small number of neurons (2 of 347) were classified as necrotic in the group of normal control subjects (Table 2⇓).
Coronal sections of the cerebral hemispheres from various experimental subgroups showed (at 2 hours) slight pallor of the lateral caudoputamen and patchy pallor of the ipsilateral cerebral cortex (Fig 3A⇓). At 4 to 6 hours, multiple foci of different staining intensity became visible within the territory of the occluded artery (Fig 3B⇓). Grossly visible swelling (as suggested by compression of the ventricles) and marked pallor of the MCA territory peaked at 72 hours, as previously reported7 ; after 5 to 6 days, swelling had subsided and coagulation necrosis became clearly visible, but this abnormality involved only the preoptic area and a fraction of the putamen (Fig 3C⇓ and 3D⇓).
Acute ischemic changes (shrinkage/scalloping and swelling) of neuronal perikarya were visible in increasing numbers at all three locations (preoptic area, caudoputamen, and cerebral cortex) as a function of time elapsed after MCA occlusion. The percentage of neurons with acute ischemic changes (shrinkage and swelling) of the type shown in Figs 1C⇑ and 1D⇑ and 2C and 2D increased from 29% (at 30 minutes) to 79% (at 6 hours) (Table 3⇓, Fig 4⇓).
Necrotic neurons (Figs 1E⇑, 1F⇑, 2E⇑, and 2F⇑) were first detected in small numbers in the preoptic area at 1 hour, but the number of these necrotic neurons did not increase significantly (65%) until 6 to 12 hours after MCA occlusion (Table 3⇑, Fig 4⇑).
The percentage of necrotic neurons visible in the frontoparietal cortex was lower than in the other two areas (preoptic area and caudoputamen) of the ischemic territory at all time points of 12 hours or less (Fig 5⇓). As late as 12 hours after MCA occlusion only 35% of the cortical neurons were classified as necrotic compared with the very high proportion of these cells visible in the caudoputamen (86%) and the preoptic area (73%) (Table 3⇑).
Pannecrosis (signs of irreversible cell injury involving all cell types, including neurons, glia, and most microvessels) was first observed in the preoptic area, striatum, and portions of the cortex 72 to 96 hours after MCA occlusion (Fig 3⇑).
Cavitation (areas of reabsorption of the necrotic debris by macrophages) was first seen in the preoptic area 5 days after MCA occlusion. After 7 days, small, isolated foci of cavitation were visible in the preoptic area and striatum. Six to 7 days after MCA occlusion a few brains also showed very small areas of cavitation at the corticomedullary junction. The time-dependent changes observed in the growth of the area of pallor and the percentage of necrotic neurons per microscopic field in each experimental subgroup are shown in Fig 6⇓.
The percentage of dark neurons in the control and experimental subgroups was the same; in addition, there were no significant differences in the percentage of dark neurons between right and left hemispheres. The first significant number of necrotic neurons was detected at 6 hours (P<.05); a second significant increase occurred at 12 hours (P<.0001). As late as 24 hours after MCA occlusion, significant differences persisted in the number of necrotic neurons between the cortex and the striatum/preoptic area (P<.05).
Dark neurons have been associated with the handling of the brain (fixation and removal)14 that is necessary for histological preparation. These experiments confirm the inference that this type of neuronal alteration is not influenced by the biological events initiated by MCA occlusion; dark neurons appeared in control subjects and in the hemisphere opposite the side of the arterial occlusion.
In these experiments we confirm that neuronal changes induced by MCA occlusion are of two types: acute changes (shrinkage and swelling) are prominent during the first 6 hours, whereas delayed changes (necrosis) affect large numbers of neurons only from 6 to 12 hours. Both types of cellular changes are time dependent, but at all time intervals of 24 hours or less, the percentage of necrotic neurons is lower in the cortex than in other territories of the ischemic hemisphere (Fig 5⇑). If one accepts the premise that the neurological deficit associated with MCA occlusion primarily reflects the extent of necrosis involving cortical neurons,15 it appears that there is a brief period (a duration of approximately 2 to 3 hours in the rat) during which interventions such as thrombolysis and reperfusion may salvage a number of cortical neurons that otherwise would be destined to die several hours later.
The ultrastructural alterations observed in neuronal perikarya in the acute stage, such as clumping of heterochromatin, dilatation of endoplasmic reticulum cisternae, and marked swelling of the mitochondrial inner matrix (Fig 2C⇑ and 2D⇑), are all considered potentially reversible on restoration of the normal blood flow.11 12 13
Cell death resulting from environmental perturbations such as altered blood flow is considered unique and is known as coagulation necrosis.11 Coagulation necrosis secondary to the occlusion of the hepatic artery is expressed in well-characterized histological abnormalities that include first pyknosis/eosinophilia and subsequently karyorrhexis and karyolysis.10 11 The biochemical mechanisms operative in coagulation necrosis include accelerated degradation of plasma membrane phospholipids and irreversible mitochondrial damage, usually accompanied by a rapid influx of Ca2+ into both the cytosol and the mitochondrial matrix.10 11 12 13
Neuronal changes of a delayed type (pyknosis/eosinophilia and loss of hematoxylinophilia) are reliable indicators of irreversible cell injury. At the ultrastructural level these delayed alterations include breaks in the plasma and nuclear membranes and deposition of electron-dense precipitates in the mitochondrial inner matrix (Fig 2E⇑ and 2F⇑).11 12 13 The mitochondrial flocculent densities represent deposits of calcium-rich salts, and excessive intracellular Ca2+ is one of the common end points of many types of irreversible cell injury, including ischemia.13
If neuronal necrosis secondary to a single-artery occlusion proceeds at a different pace in various regions of the ischemic territory, could the development of necrosis at a given site be the reflection of spatial and time-dependent differences in local blood flow values?
Data on the nature of the blood flow changes induced by permanent occlusion in this paradigm of brain infarct are incomplete.16 17 Nagasawa and Kogure17 calculated terminal CBF values after injecting [14C]iodoantipyrine. They reported a time-dependent progressive decrease in CBF values: 3- and 6-hour occlusions were worse than 1-hour MCA occlusion. Moreover, at all time intervals CBF values were lower in the caudoputamen than in the cortex.17 These observations correlate well with the progression of the lesion as evaluated by counts of necrotic neurons; the combined results of the two experiments suggest that as local CBF values deteriorate, an increasing number of neurons become necrotic.
Local CBF values were calculated by the [14C]iodoantipyrine method in a study of transient MCA occlusion, induced by the same method used in our experiments, in six rats subjected to 60-minute MCA occlusion and 15-minute recirculation.18 Reperfusion of the previously ischemic structures was noted in all rats. Local CBF values ranged from 19% (frontoparietal cortex) to 47% (lateral caudoputamen) of the contralateral values.18 The study aimed at estimating the period of reversibility for this type of arterial occlusion, but it left incompletely answered these questions: (1) Is there a progressive worsening of the CBF at different sites of the MCA territory? (2) Does reopening the artery (within a reasonable period of time) decrease the number of necrotic neurons and improve neurological function?
Schmid-Schönbein and Engler19 suggested that after a coronary occlusion, the lumina of numerous microvessels in the territory of the occluded artery become obstructed by polymorphonuclear leukocytes, which could create a significant impediment to the circulation of erythrocytes. The progressive interference with erythrocyte circulation could be one of the factors responsible for neuronal necrosis. In the paradigm of brain infarct we used, the influx of leukocytes begins very early (30 minutes after MCA occlusion), but peak numbers of intravascular leukocytes are detected only 12 hours later.20 This is the same time when large numbers of necrotic neurons become detectable; however, factors responsible for leukocyte migration and the role that these cells may play in the evolution of the brain lesion remain incompletely defined. Mori et al21 tested the hypothesis that inhibiting leukocyte/endothelial cell adhesion with a monoclonal antibody against the CD11b surface receptor would improve the patency of brain microvessels in baboons with transient MCA occlusion. Significant improvement was obtained in the animals receiving the antiserum compared with those given a placebo (P<.034 to .049). Working with the paradigm of brain injury secondary to intravascular MCA occlusion in rats, Chen et al22 23 24 reported a decrease in the area of brain pallor by a combination of arterial reopening (2 hours after MCA occlusion) and administration of the anti-CD11b monoclonal antibody. However, no comments were made in these publications about the effects on number of necrotic neurons. Several authors have reported successful intervention, as measured by improvements in the volume of the area of pallor, among animals with MCA occlusion that receive various therapeutic compounds.16 22 23 24 25 26 However, experiments demonstrating a validated correlation between degree of neurological deficit and volume of area of pallor are not available. This may explain the discrepancy noted by Wiebers et al27 between the success reported in many animal experiments aimed at improving the effects of MCA occlusion28 and the lack of beneficial effects noted in clinical trials in which similar compounds have been administered to patients with ischemic stroke.
This study was supported in part by US Public Health Service grant NS-31631 (Dr Garcia). The authors thank Jun Xu for the surgical preparations necessary to conduct these experiments; Lisa Pietrantoni, registered HTL (ASCP), for the histological and electron microscopy preparations; and Paula McGee and Kathy Zajas for expert secretarial support.
- Received August 19, 1994.
- Revision received November 7, 1994.
- Accepted December 29, 1994.
- Copyright © 1995 by American Heart Association
- ↵Astrup J, Symon L, Siesjö BK. Thresholds in cerebral ischemia: the ischemic penumbra. Stroke. 1981;12:723-725.
- ↵Heiss WD. Experimental evidence of ischemic thresholds and functional recovery. Stroke. 1992;23:1668-1672.
- ↵Zea Longa E, Weinstein PR, Carlson S, Cummins R. Reversible middle cerebral artery occlusion without craniectomy in rats. Stroke. 1989;20:84-91.
- ↵Animal welfare: final rules (9CFR parts 1, 2, and 3). Federal Register. August 31, 1989;54:36112-36163. US Dept of Agriculture, Animal and Plant Health Inspection Service.
- ↵Trump BF. Significance of mitochondrial conformational changes in injured cells. Am J Pathol. 1973;70:83A. Abstract.
- ↵Cammermeyer J. Nonspecific changes of the central nervous system in normal and experimental material. In: Bourne GH, ed. The Structure and Function of the Nervous Tissue. New York, NY: Academic Press, Inc; 1972: vol 6.
- ↵Garcia JH, Wagner S, Liu K-F, Hu X-J. Neurological deficit and extent of neuronal necrosis attributable to middle cerebral artery occlusion: statistical validation. Stroke. 1995;26:627-635.
- ↵Kawamura S, Shirasawa M, Fukasawa H, Yasui N. Attenuated neuropathology by nivaldipine after middle cerebral artery occlusion in rats. Stroke. 1991;22:51-55.
- ↵Nagasawa H, Kogure K. Correlation between cerebral blood flow and histologic changes in a new rat model of middle cerebral artery occlusion. Stroke. 1989;20:1037-1043.
- ↵Memezawa H, Smith ML, Siesjö BK. Penumbral tissues salvaged by reperfusion following middle cerebral artery occlusion in rats. Stroke. 1992;23:552-559.
- ↵Mori E, del Zoppo GJ, Chambers D, Copeland BR, Arfors KE. Inhibition of polymorphonuclear leukocyte adherence suppresses no-reflow after focal cerebral ischemia in baboons. Stroke. 1992;23:712-718.
- ↵Chen H, Chopp M, Bodzin G. Neutropenia reduces the volume of cerebral infarct after transient middle cerebral artery occlusion in the rat. Neurosci Res Commun. 1992;11:93-99.
- ↵Chopp M, Zhang RL, Chen H, Li Y, Jiang N, Rusche JR. Postischemic administration of an anti–Mac-1 antibody reduces cell damage after transient middle cerebral artery occlusion in rats. Stroke. 1994;25:869-876.
- ↵Bielenberg GW, Burniol M, Rosen R, Klaus W. Effects of nimodipine on infarct size and cerebral acidosis after middle cerebral artery occlusion in the rat. Stroke. 1990;21(suppl IV):IV-90-IV-92.
- ↵Wiebers DO, Adams HP Jr, Whisnant JP. Animal models of stroke: are they relevant to human disease? Stroke. 1990;21:1-3.